ORGANIZING PROCESSES AND THE CONSTRUCTION OF …...Smallman, 1996). Despite some positive...

ORGANIZING PROCESSES AND THE CONSTRUCTION OFRISK: A DISCURSIVE APPROACH

STEVE MAGUIREMcGill University

CYNTHIA HARDYUniversity of Melbourne

This study examines the organizing processes through which products “become” risky.Drawing on a case study of chemical risk assessment and management processes inCanada and comparing two chemicals, it identifies a series of enacted practices thatbundle into two forms of social ordering: “normalizing” and “problematizing.” Bybringing the past to bear differently on organizing processes, these two forms of socialordering structure the discursive work of actors in both their attempts to stabilize andtheir attempts to destabilize and change meanings of risk objects. As a result, objects“become” risky or safe in different ways.

This study examines organizing processesthrough which products “become” risky—or, forthat matter, safe. The processes whereby productsand technologies are deemed to pose risks are im-portant in our global risk society (Beck, 1992). Al-though risk can be viewed positively, increasinglyit is associated less with opportunities for gain andmore with possibilities of loss (Gephart, VanMaanen, & Oberlechner, 2009). Accordingly, risk istypically defined as “the potential for realization ofunwanted, adverse consequences to human life,health, property, or the environment” (Society forRisk Analysis [SRA], 2003). Organizations have tomanage risk (Power, 2007; Smallman, 1996) anddeal with the consequences when their productsand technologies are found to pose risks to externalstakeholders (Maguire, 2004; Scheytt, Soin, Sahlin-Anderson, & Power, 2006) Organizations are alsoplaces where risks are conceptualized, measured,and assessed: they are “centres for processing andhandling risk” (Hutter & Power, 2005: 1). It is im-portant, therefore, that we learn more about theorganizing processes whereby products become as-sociated with risks and steps are taken to man-age them.

Our study examines chemical risk assessmentand management processes in Canada. In 2006, theCanadian Government completed the task of cate-

gorizing over 23,000 existing substances andlaunched a follow-up process to assess whether 200“high priority” chemicals were toxic—that is,whether they posed risks to human health and theenvironment. Some chemicals that had long beenconsidered safe “became” risky, requiring govern-ment action to reduce or eliminate (i.e., manage)the risks they posed, while others remained safe.We compare the fate of two chemicals during theseprocesses: vinyl acetate monomer (VAM) and bis-phenol A (BPA). VAM is used to make a variety ofindustrial and consumer products. It was found tobe toxic in an initial assessment; but this findingwas subsequently reversed, and the Governmentdeclared that VAM was safe. BPA is used in a rangeof consumer products including baby and waterbottles. It was found to be toxic, as a result of whichCanada became the first country in the world toimplement a ban on baby bottles containing thischemical. The meaning of these products—risky orsafe—is thus temporally and spatially contingent: itchanges over time and from place to place. VAMhad been considered safe in the past; appeared tobe on course for being considered risky in Canada;but then was again declared safe. BPA had alsobeen considered safe in the past; was still consid-ered safe in most countries; but was considered anunacceptable risk in Canada for use in baby bottles.These chemicals are, accordingly, in an ongoingstate of becoming and “unbecoming” risky.

We use a performative process perspectivewherein social phenomena are in a perpetual stateof becoming (Tsoukas & Chia, 2002) to understandhow the meanings of VAM and BPA as “risk ob-jects” (Hilgartner, 1992) were constructed. We com-

The authors gratefully acknowledge the financial sup-port of the Social Sciences and Humanities ResearchCouncil of Canada (SSHRC award number 410-2011-1265) and the Australian Research Council (Discoveryfunding scheme, project number DP 0771639). They alsothank Adam Halpert for his invaluable researchassistance.

� Academy of Management Journal2013, Vol. 56, No. 1, 231–255.http://dx.doi.org/10.5465/amj.2010.0714

231

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s expresswritten permission. Users may print, download, or email articles for individual use only.

bine this perspective with a discursive approach toexamine the discursive work; (producing, distrib-uting, and consuming texts) that occurred aroundthese meanings. Our findings differentiate twoforms of social ordering (Chia, 2002)—normalizingand problematizing—relevant to contemporary riskassessment and management organizing processesand show how they structured the discursive workof actors in shaping the meanings of these chemi-cals. Our study makes the following contributions:First, it illustrates empirically how risks emergefrom the very organizing processes through whichthey are assessed and managed: objects “become”risky or safe in different ways as a result of thepractices to which they are subjected. Second, itshows how these practices bundle into differentforms of social ordering, which structure the dis-cursive work of actors in both their attempts tostabilize and their attempts to destabilize andchange meanings of risk objects. Third, it showshow meanings are constructed at the intersection ofpractices and discursive work and how both prac-tices and discursive work are required to holdmeanings in place.

RISK, ORGANIZING PROCESSES,AND DISCOURSE

Risk is a prominent feature of contemporary or-ganizations and their environments: the productionof risk has become as important as that of wealth(Tsoukas, 1999), and managers are increasinglyhaving to engage in risk management (Power, 2007;Smallman, 1996). Despite some positive associa-tions, such as taking advantage of opportunities,risk is increasingly likely to refer to “negative andundesirable consequences” (Gephart et al., 2009:141), the “anticipation of catastrophe” (Beck, 2006:332), or “the chance of mishap” (Cranor, 2007: 38).In what has been termed the risk society (Beck,1992), risks have become “more global, less readilyidentifiable, more problematic, less easily man-aged, and more anxiety-provoking” (Gephart et al.,2009: 142); social institutions are increasingly pre-occupied with risks but, paradoxically, less able tomanage them (Mythen, 2008; Tsoukas, 1999). Dis-cussions of organizational side effects (Scheytt etal., 2006) emphasize that risks are unavoidable con-sequences of organizing; are often borne by a widerange of actors outside risk-producing organiza-tions; and, therefore, lead to greater scrutiny andregulation of organizations’ activities (Eccles,Newquist, & Schatz, 2007; Power, 2005: 2007). Con-sequently, the processes whereby products andtechnologies are assessed as posing risks—and howthose risks are subsequently managed—are of con-

siderable significance to organizational scholars,practitioners, and the public alike.

Theorizing by natural scientists has given rise toa well-developed scientific discipline, risk analy-sis, from which has emerged an established set ofpractices for assessing, managing, and communi-cating risks (SRA, 2003). Much of the research onrisk is, as a result, dominated by perspectives fromthe natural sciences, engineering, and medicine(Miller, 2009) with a “realist” approach (Jasanoff,1998). An assumption in this research is that risksare “objectively quantifiable” phenomena (Miller,2009: 30) that scientific experts can identify andmeasure as “the expected value of the conditionalprobability of the event occurring times the conse-quence of the event given that it has occurred”(SRA, 2003). The realist approach is integral tounderstanding risk. Its assumptions form the basisof the vast majority of formal risk assessment andmanagement processes that help decision makersto estimate harms associated with possible coursesof action (Renn, 1992, 2008). Technical risk analy-sis “is vital for making risk decisions better in-formed, more consistent, and more accountable”(Kunreuther & Slovic, 1996: 123). Moreover, man-agers tend to act as if risks are—or at least shouldbe—measurable and assessable (e.g., Sullivan-Tay-lor & Wilson, 2009).

The realist approach has, however, been criti-cized. First, critics argue that it privileges scientificknowledge at a time when, instead of creating morecertainty, as once “triumphantly presumed, scienceincreasingly generates even more uncertainty”(Tsoukas, 1999: 505). Second, not all risks can bequantified and, for those that can, different mea-surements may result in quite different assessments(Kunreuther & Slovic, 1996). Third, critics main-tain that value judgments are inherent in risk as-sessments (Pidgeon, Hood, Jones, Turner, & Gibson,1992) and therefore need to be acknowledged(Miller, 2009; Renn, 2008). Fourth, the realist ap-proach tends to dismiss lay understandings of riskas a distortion of “actual” risk as defined by experts(Gephart et al., 2009; Jasanoff, 1998). Yet publicperceptions are frequently incorporated into riskassessment and management (Kunreuther & Slovic,1996), and scholars argue that such “extrascien-tific” knowledge (Wynne, 1992: 294) should be“considered not as error but as an essential datum”(Pidgeon et al., 1992: 91).

As a result, an alternative approach, involvingconstructionist accounts of risk, has emerged toexplore risk as a social phenomenon (Lupton, 1999;Renn, 2008; Zinn, 2008). Its argument is that risksare never “fully objective or knowable outside ofbelief systems and moral positions” (Gephart et al.,

232 FebruaryAcademy of Management Journal

2009: 144). They do not exist “out there,” “waitingto be measured” (Kunreuther & Slovic, 1996: 119)but are, instead, subjective. This approach ad-vances the idea that “the fundamental doctrines ofrisk management are inherently plural, disputableand disputed” (Hood & Jones, 1996: xi–xii). Multi-ple interpretations of what poses a danger are pos-sible because of the inherent ambiguity of socialphenomena (Tansey & O’Riordan, 1999), with theworldviews of different actors serving as lenses,“magnifying one danger, obscuring another threat,selecting others for minimal attention” (Dake, 1992:33). In other words, risks are not discovered but are“defined, perceived and managed according toprinciples that inhere in particular forms of socialorganization” (Raynor, 1992: 849).

We build on this constructionist approach to ex-amine the way in which risk and organizing pro-cesses implicate each other (cf. Malenfant, 2009;Miller, 2009; Winch & Maytorena, 2009). We areinterested in how risk is “socially constructedthrough varied processes of negotiation and con-flict resolution in settings ranging from the rela-tively closed quarters of a research laboratory to thepublic debate of a regulatory hearing” (Jasanoff,1998: 94). According to Hilgartner (1992: 40), thisprocess involves the construction of “at least threeconceptual elements: an object deemed to ‘pose’the risk, a putative harm, and a linkage allegingsome form of causation between the object and theharm.” This is not to deny that individuals facepotential harms from some objects; that technicalanalyses can illuminate these harms; or that theseanalyses involve scientific practices that are basedon a realist understanding of risk. Rather, it focusesattention on the specific ways in which particularmeanings are attached to objects through organiz-ing processes that construct risks, as well as on howrisks are shaped by the very processes used to as-sess and manage them.

Organizing Processes and Risk

To explore how risks and the meaning of riskobjects are constructed in organizational settings,we draw on a process approach to organizing (Vande Ven & Poole, 2005). Also known as “organiza-tional becoming” (Chia, 1995; Nayak, 2008; Tsou-kas & Chia, 2002), these performative accounts oforganizing processes collapse traditional ontologi-cal distinctions between process and object. Ac-cording to this view, objects are not preexisting orfixed but, rather, are temporary patterns “consti-tuted by and shaped from micro-interactions asactors perform their everyday work” (Thomas, Sar-gent, & Hardy, 2011: 22). They are “products of

processes rather than existing prior to them” (Bak-ken & Hernes, 2006: 1600; emphasis added). Cer-tain objects may appear to be stable and immutable(Carlsen, 2006), but this is only because their mean-ing is held in place through countless communica-tive interactions among actors (Tsoukas &Chia, 2002).

The objects that humans think they perceive are,then, abstractions that emerge from organizing pro-cesses through which particular meanings are at-tributed. The forms that these abstractions take—the objects humans think they see—are notcompletely open-ended or random; they are in-formed by and structured through practices that areenacted at a given moment in time but also emanatefrom the past. Distinct practices are bundled inso-far as they become “coordinated or linked with oneanother and also exhibit temporal features such asrhythm and patterning” (Schatzki, 2006: 1866) asthey unfold over time. Accordingly, disparate prac-tices can attain a degree of coherence as possiblecourses of action present themselves and somecombinations are selected over others (Bakken &Hernes, 2006).

Chia (2002) referred to persistent patterns amongpractices as forms of social ordering that emerge asorganizations establish rhythms that become frame-works for regulating interactions and modes ofthought.

Such forms of social ordering inevitably influence,amongst other things, how the flux and flow of ourlife-worlds are structured and conceptualized intoevents, things and situations; how identity is estab-lished and social entities created; how taxonomiesand systems of classifications are produced andwith what effects; how reification takes place, howcausal relations are imputed, and with what conse-quences. (Chia, 2002: 867)

A form of social ordering can thus be defined as abundle of interrelated practices that persists overtime. This “temporalizing” (Hodges, 2008: 409)brings the past to bear on the present by connectingobjects to wider social structures and brings thepast to bear on the future by enabling and con-straining outcomes. Without such a “pre-structuredfield of possibilities,” it would not be possible “toestablish the identity of an object of analysis”(Chia, 1999: 219).

In sum, objects are both “momentary outcomes”of localized, situated practices and “effects of his-torical processes” (Chia, 2002: 866) that emergefrom the patterned enactment of practices as actorsengage in organizing processes over time. In apply-ing this approach, one must therefore examine how“bundled activities interweave with ordered con-

2013 233Maguire and Hardy

stellations of non-human entities” (Schatzki, 2001:3) or objects. According to this perspective, then, arisk object is a “physically stabilized, congealedembodiment” (Jasanoff, 1998: 97) of past social as-sumptions and conventions, rather than a preexist-ing object with essentially harmful characteristics.Anything can be a risk. “It all depends upon howone analyzes the danger, considers the event”(Ewald, 1991: 199). Precisely how dangers are an-alyzed and events are considered depend upon theparticular practices in the organizing processesthrough which risk is assessed and managed. Ac-cordingly, our first research question is: How arethe meanings of risk objects constructed throughorganizing processes?

Discursive Work and Risk

Although organizing processes play a role in con-structing the meaning of risk objects so, too, doesdiscourse (Tsoukas, 2005). Practices are performedwithin the context of discourses—structured col-lections of texts and discursive practices (Phillips,Lawrence, & Hardy, 2004) that define “who andwhat is ‘normal,’ standard and acceptable” (Meri-läinen, Tienari, Thomas, & Davies, 2004: 544) andposition actors in such a way that not all warrantvoice (Hardy & Phillips, 1999). Discourse con-strains “the strategies and rules by which we canspeak about and act on a domain of objects . . . insuch a way that certain possibilities and outcomesare realized rather than others” (Reed, 1998: 196).Accordingly, when organizing processes occur inthe context of a dominant discourse, meanings tendto be stabilized in ways that have become taken-for-granted (e.g., Maguire & Hardy, 2009; Phillips etal., 2004; Robichaud, Giroux, & Taylor, 2004). Inthe case of risk, the pervasiveness of realist ap-proaches has led to a dominance of a discourse thatprivileges formal, scientific knowledge, empower-ing some people as experts and dismissing othersas inexpert.

The discursive model [of risk] also views knowledgeabout risk as socially constructed, but it emphasizesthe role of professional languages (such as quantita-tive risk assessment) and analytic practices (such ascost-benefit analysis) in shaping public perceptionsof risk. Authoritative knowledge is created in thisframework by people or institutions that master therelevant formal discourses, which, however, impor-tantly constrain even the experts’ perceptions ofrisk. (Jasanoff, 1998: 94)

Discourses are, however, never completely cohe-sive: they are always partial, inconsistent, and con-tested to some degree and, therefore, never able to

determine social reality fully (Hardy & Phillips,2004). As a result, actors can also draw on them tochallenge what was once taken-for-granted in orderto destabilize meanings, thereby forming a basis ofchange (e.g., Heracleous & Barrett, 2001; Tsoukas,2005) through their discursive work—that is,through their producing, distributing, and consum-ing texts (Hardy & Maguire, 2010; Hardy & Phillips,2004; Phillips et al., 2004).

In the context of risk, “the process of construct-ing a risk object” is a “rhetorical process performedin texts” (Hilgartner, 1992: 46). A discursive per-spective on organizing processes can therefore“highlight the ways in which dominant meaningsemerge” as well as shed light on “the discursivepractices and rhetorical devices that are deployed”in “struggles around meaning” (Grant & Hardy,2004: 5). Accordingly, our second research ques-tion is: What is the role of discursive work in sta-bilizing (and destabilizing) meanings of riskobjects?

METHODS

Site Selection

Our study examines processes of assessing andmanaging the risks of chemicals in Canada. Weselected Canada because it was “the first country inthe world to categorize the thousands of chemicalsubstances in use before comprehensive environ-mental protection laws were created” (Governmentof Canada, 2007a). There is, therefore, much tolearn that may be relevant for other countries. Sec-ond, these processes took place over a number ofyears, making it possible to examine how meaningschanged over time. Third, Canada had publishedmany texts that we could analyze. Fourth, one ofthe authors has a formal role in the process and isfamiliar with the context.1 We compare the fate oftwo chemicals—vinyl acetate monomer (VAM) andbisphenol A (BPA)—because their meaningsclearly changed over time.

Case Study

The 1999 Canadian Environmental ProtectionAct (CEPA) put in place a formal process for thesystematic categorization of 23,000 chemicals al-ready in use according to the risks they could poseto human health and the environment. Figure 1

1 One of the authors is a member of the ChallengeAdvisory Panel (see below). Please note that the opinionsexpressed here reflect those of the authors and not thoseof the Challenge Advisory Panel.


summarizes these chemical risk assessment andmanagement processes in Canada. The Governmentidentified 200 products as “high priorities for ac-tion” (Government of Canada, 2010a) and listedthem as part of what is called “the Challenge,”which meant that it was “predisposed to treat”them as “CEPA-toxic”—toxic according to the cri-teria stated in the 1999 Act.2 Industry and otherstakeholders were then “challenged” to submit spe-cific information within certain timelines to informthe risk assessment and management processes. Inthe Challenge, assessing whether a substance isCEPA-toxic is guided by a weight of evidence ap-proach and the precautionary principle (Environ-ment Canada, 2005). The former involves integrat-

ing multiple measures or bodies of evidence fromprior studies, taking into consideration theirstrengths and weaknesses. The latter affirms that“lack of full scientific certainty shall not be used asa reason for postponing cost-effective measures toprotect the environment and human health” (Gov-ernment of Canada, 2009a).

The Challenge is managed by two Governmentdepartments, Health Canada and Environment Can-ada, working together to assess 12 batches of 12–20chemicals according to strict timelines over a pe-riod covering several years. For each chemical in abatch, stakeholders are invited to provide informa-tion, following which a draft “screening assessmentreport” (SAR) is produced. Stakeholders are theninvited to comment on the draft SAR. If a draft SARproposes that a chemical is toxic, then a “risk man-agement scope” (RMS) document is produced sum-marizing the issues and inviting stakeholders tosubmit more information. Stakeholders’ commentsare summarized in the “summary of public com-ments” (SPC) and factored into the final SAR,which may or may not confirm the original conclu-sion. If the final SAR concludes that a chemical isnot CEPA-toxic, no further action is taken. If achemical is found to be toxic, it may be added to

2 According to Article 64 of CEPA (1999), “a substanceis toxic if it is entering or may enter the environment ina quantity or concentration or under conditions that (a)have or may have an immediate or long-term harmfuleffect on the environment or its biological diversity; (b)constitute or may constitute a danger to the environ-ment on which life depends; or (c) constitute or mayconstitute a danger in Canada to human life or health”(http://www.ec.gc.ca/lcpe-cepa/documents/archives/loi-act/CEPA_full-eng.pdf, accessed July 2010).

FIGURE 1Chemical Risk Assessment and Management Processes in Canada

1999

2009 December

2006

2006December

2007 2007 2007 February

23,000

chemicals categorized; 4,000 require further action

Batch 1 of Challenge announced

Batch 2 of Challenge announced; it includes BPA, VAM

“Challenge” announced for 200 highpriority chemicals

May

August

Batch 3 of Challenge announced

Final Batch 12 of Challenge announced

CEPA 1999 requires categorization of all existing chemical substances

Public comment period

Draft assessment designates VAM as toxic

Draft assessment designates BPA as toxic


Final assessment designates BPA as toxic; risk management approach is proposed

Final assessment designates VAM as not toxic


BPA added to List of Toxic Substances

Advisory Panel is consulted and agrees with Government approach

Ongoing risk management measures, including ban on baby bottles containing BPA

Advisory Panel is consulted and agrees with Government approach

SEQUENCE OF EVENTS FOR VAM AND BPA

VAM No further measures

Information is submitted by stakeholders

Information is submitted by stakeholders

BPA


the “Priority Substances List” for further assess-ment of its risks, or to the “Toxic Substances List,”in which case the Government must prepare a “pro-posed risk management approach” (PRMA) to ad-dress the risks. Stakeholders are provided with op-portunities to comment on these proposals,following which a range of measures, such as reg-ulations, guidelines, or codes of practice, can beimplemented “to control a chemical’s research anddevelopment, manufacture, use, storage, transportand/or ultimate disposal” (Government of Can-ada, 2010b).

During these processes, information on a chemi-cal’s properties, hazards, uses, releases to the envi-ronment, and routes of exposure to humans is col-lected and analyzed. This work is undertaken bystaff at Health Canada and Environment Canadaand peer-reviewed by external scientists. The Gov-ernment also solicits and receives advice on spe-cific questions from the Challenge Advisory Panel(CAP), which has 12 members selected for theirknowledge and expertise (not as organizational orprofessional representatives).

Vinyl acetate monomer (VAM). VAM is an in-dustrial chemical used to manufacture a wide va-riety of polymers and is found in paints, adhesives,and personal care products in the form of residuefrom the manufacturing process. Scientific findingshad suggested that long-term exposure to vinyl ac-etate can cause a carcinogenic response (Vinyl Ac-etate Council, 2009). VAM was therefore includedin the Challenge. The draft SAR, released in May2008, stated that VAM was toxic to humans on thebasis of an assumption that it is a “nonthreshold”carcinogen, which is one that may harm humanhealth at any level of exposure rather than only atexposure above a threshold level (EnvironmentCanada and Health Canada, 2008: 2). This was fol-lowed by a 60-day public comment period duringwhich 28 submissions were received from industryorganizations, individual companies, and nongov-ernmental organizations (NGOs), many of whichcommented on VAM’s mode of action (i.e., whetherit was nonthreshold or threshold).

The Government subsequently accepted thatVAM had a threshold mode of action and, drawingon data on exposure scenarios, concluded that itdid not meet the criteria for toxicity. In doing so, itconsulted the Challenge Advisory Panel in October2008 on whether it agreed “that the weight of evi-dence and application of precaution” supportedthis conclusion (VAM-CAP, 2008). The Panelagreed. Accordingly, the final SAR reversed thefinding of the draft SAR and, instead, concludedthat VAM was not toxic. As a result, a PRMA doc-ument was not required, and the Challenge file on

VAM was closed with a summary of these pro-cesses posted on the website.

Bisphenol A (BPA). BPA is used in baby andwater bottles, sports equipment, lenses, CDs andDVDs. Indications of possible health risks emergedin the 1990s, as scientific findings suggested it wasan endocrine disruptor (i.e., a chemical that inter-feres with hormones to cause reproductive, devel-opmental, neural, or other problems). In 2006, itbecame part of the Challenge and underwent as-sessment. In March 2008, the Government askedthe Challenge Advisory Panel to comment on theGovernment’s approach to BPA. The Panel agreedthat “the weight of evidence and the application ofprecaution support the conclusion reached byHealth Canada”—that is, that BPA was toxic (BPA-CAP, 2008). In April 2008, the draft SAR stated thatBPA was toxic to both humans and the environ-ment and proposed that it be added to the ToxicSubstances List. A 60-day public comment periodduring which 21 submissions were received fromindustry organizations, NGOs, public health organ-izations, and individuals followed (Government ofCanada, 2009b). In October, the final SAR con-firmed that BPA constituted a potential danger inthe form of neurological effects in early stages ofchild development (Government of Canada,2009b). The PRMA, published the same month,outlined a proposed ban on baby bottles containingBPA, initiating another 60-day public comment pe-riod, which saw the receipt of 15 submissions (Gov-ernment of Canada, 2009b). In June 2009, it wasannounced that BPA would be added to Schedule Iof the Hazardous Products Act (Government ofCanada, 2009b) to allow for the prohibition of theadvertisement, sale, and importation of baby bot-tles containing BPA. In March 2010, this ban wasimplemented.

The ban on baby bottles did not close the file onBPA since the Government continues to carry outactivities related to its potential risks. As of August2010, Environment Canada was monitoring BPAand studying landfills to examine the life cycle ofproducts containing BPA (Government of Canada,2010c), while Health Canada was working with thefood packaging industry on the implementation ofvoluntary measures to reduce levels of BPA in in-fant formula products (Government of Can-ada, 2009c).

Data Collection

We collected publicly available texts that de-scribed the practices of chemical risk assessmentand management processes, starting with theCEPA. We then searched the Government’s web


portal on chemical substances to identify anddownload texts on the overall processes (www.chemicalsubstanceschimiques.gc.ca) as well as alldocuments pertaining to VAM and BPA. Wesearched the websites of Environment Canada andHealth Canada to find texts related to risk assess-ment and management processes and to the twochemicals. (The Appendix cross-references officialChallenge documents addressing the two chemi-cals with our in-text citing notation, to make iteasier for readers to make connections between par-ticular quotations and specific types of texts.) Wealso collected publicly available documents pro-duced by NGOs and industry relating to the assess-ment of VAM and BPA by searching the internet.

Data Analysis

In the first stage of analysis, we constructed anevent history database (Van de Ven & Poole, 1990)that captured what happened when. We also createda discursive event history database (Maguire, 2004)by ordering texts chronologically, identifying author-ship, and classifying content in relation to VAM andBPA. From these databases, we developed an overalltimeline of the risk assessment and management pro-cess and, in particular, what happened with VAMand BPA. This timeline was used as the basis of thecase study description above.

In the second stage of analysis, we identified textsin our database that described chemical risk assess-ment and management processes in Canada. Using aninductive, interpretive approach, we systematicallycoded these descriptions, from which we inferredeight distinct practices related to risk assessment and

management. We refer to these practices as “referenc-ing,” “anchoring,” “categorizing,” “sequencing,”“particularizing,” “innovating,” “questioning,” and“pluralizing.” Table 1 summarizes how these prac-tices were identified. We also noted patterns amongthese practices: whether certain practices tended toco-occur and could be linked conceptually. On thebasis of this interpretive analysis we clustered thepractices into two distinct bundles of four interre-lated practices. We refer to these bundles as normal-izing and problematizing and discuss them in moredetail in the first set of findings.

In the third stage of analysis, we examined thesets of texts that were produced as part of the riskassessment and management of VAM and BPAfor traces of the eight practices. Table 2 summa-rizes the documents that we analyzed. We sys-tematically coded the two sets of texts for evi-dence of the two bundles of practices andcompared the results. We found considerable ev-idence of normalizing practices in the case ofVAM, in that the texts produced in the assess-ment of VAM contained numerous traces of thefour practices associated with normalizing andvery few traces of the four practices associatedwith problematizing. The texts produced in theassessment of BPA also contained traces of nor-malizing practices, but there was also extensiveevidence of the practices associated with prob-lematizing. The results of this stage of the analy-sis are discussed in the second set of findings.

Finally, we analyzed the discursive work carriedout around VAM and BPA. We began by comparingthe texts produced for VAM with those produced forBPA to note similarities and differences in their num-

TABLE 1Practices Identified in Canadian Chemical Risk Assessment and Management Organizing Processes

Description of Data Coded Practices

Bundles ofInterrelated

Practices

Instances in which systematic application of accepted scientific knowledge and methodsis emphasized with reference to extant research, scientific experts, and otherjurisdictions

Referencing

Instances in which current activities are related to past activities, decisions, experience,and/or precedents, and continuity is emphasized

Anchoring Normalizing

Instances in which clearly bounded categories are used, from which certain actionsfollow i.e., if X then Y

Categorizing

Instances in which actions are temporally sequenced through pre-established timelines,charts, flow diagrams, etc.

Sequencing

Instances in which a case is made for unique considerations or special treatment ParticularizingInstances in which activities are distinguished from those carried out in the past and

novelty is emphasizedInnovating

Problematizing

Instances in which uncertainty and incompleteness of information are emphasized QuestioningInstances in which the involvement of stakeholders other than scientists and government

officials in risk assessment and management processes is emphasizedPluralizing


ber, types, length, and general contents. We then ex-plored specific attempts to stabilize, as well as todestabilize and change, the meanings of the chemi-cals by examining the SPC for each chemical. Thisdocument summarized stakeholders’ disagreementswith, and criticisms of, the conclusion of the draftSAR and provided the Government’s response. Fo-cusing on the producing, distributing, and consum-ing of texts, we identified three ways in which theconclusions of the draft assessments were contested:stakeholders produced and submitted texts in whichthey (a) criticized that important texts had been ig-nored and recommended these texts be included inthe final assessment; (b) criticized particular texts

that had been referenced in the draft report and rec-ommended that these texts be excluded from the finalassessment; and (c) criticized the way in which theGovernment had (mis)interpreted particular texts.We noted two forms of response: (a) rebuffing criti-cisms and (b) accommodating criticisms. Table 3 de-scribes how we coded the data to identify these strug-gles over meaning. We then compared the two SPC’sto ascertain any differences in the patterns of criti-cisms and responses and to see whether and howthese patterns related to the discursive work carriedout around each chemical, as well as to normalizingand problematizing practices, as we discuss in thethird set of findings.

TABLE 2Time Lines and Relevant Documents for BPA and VAM

Document VAM BPA

Summary of Consultation ofChallenge Advisory Panel(CAP)

Panel consulted in October 2008 (see below). March 2008: Panel asked to consider twoquestions related to application of the weightof evidence and precaution for BPA.

Draft Screening AssessmentReport (draft SAR)

May 2008 (41 pages): Indicated the Governmentwas considering designating VAM as toxicunder CEPA.

April 2008 (106 pages): Indicated theGovernment was considering designatingBPA as toxic under CEPA.

Risk Management Scope(RMS)

May 2008 (6 pages): Summarizes draft SAR andinvites stakeholders to submit comments.

April 2008 (8 pages): Summarizes draft SARand invites stakeholders to submitcomments.

Summary of Consultation ofChallenge Advisory Panel(CAP)

October 2008: Panel asked to consider twoquestions related to application of the weightof evidence and precaution for VAM.

Panel consulted in March 2008 (see above).

Summary of PublicComments (SPC)

November 2008 (9 pages): Summarized commentsfrom 28 industry organizations and 3 NGOs.

October 2008 (26 pages): Summarizedcomments from 8 industry organizations, 16NGOs, 4 public health organizations, and 4individuals.

Final Screening AssessmentReport (final SAR)

November 2008 (47 pages): VAM does not meetany of the CEPA criteria for toxicity.

October 2008 (107 pages): BPA meets one ormore of the CEPA criteria for toxicity.

Proposed Risk ManagementApproach(PRMA)

Not applicable. October 2008 (19 pages): Proposes ban on theimportation, sale, and advertising ofpolycarbonate baby bottles made with BPA.

October–December 2008: During consultationon PRMA, 15 submissions received from 4industry associations, 8 NGOs, 1 governmentdepartment, and 2 individuals.

Order adding a toxicsubstance to the list

Not applicable. (January 2009: notice given thatGovernment intends to take no further action).

May (2009): Notice of intention to add BPA tothe list of toxic substances on grounds thatthe assessment has found it to be toxic.

Notice of amendment toSchedule 1 of HazardousProducts Act

Not applicable. June 2009: Government announces that itproposes to add polycarbonate baby bottlesthat contain BPA to Schedule 1 to allow forprohibiting of advertisement, sale, andimportation of these products.

Final notice Not applicable. October 2009: BPA added to the list of toxicsubstances.

Web-based summary Reports that exposure levels are not consideredharmful to human health.

Reports that BPA has been determined to betoxic and that Government is proposingmeasures.

Web-based fact sheet None found. Provides further information on riskmanagement and advice for consumers.

Other documents None found. Various additional documents found on thewebsite, e.g., Government press releases onBPA.


FINDINGS

We present three sets of findings: practices relevantto Canadian chemical risk assessment and manage-ment processes; the particular practices enacted inthe construction of VAM and BPA as risk objects; andthe discursive work carried out in struggles over themeanings of these two these chemicals.

Practices Relevant to Chemical Risk Assessmentand Management Processes in Canada

In this section, we describe two bundles of inter-related practices, normalizing and problematizing.

Normalizing. Normalizing refers to an interre-lated set of practices that emphasize “normal” sci-ence. It involved the mindful application of ac-

TABLE 3Coding Categories for Struggles over Meaning

Illustration of Data Coded Description Coding Category

Criticisms of the Draft SAR“The public comments cited several limitations for this

study and recommended an alternative chronic LO(A)EC [lowest observed adverse effect concentration] of704 mg/m3 (200 ppm) based on a study in mice andrats (Bogdanffy et al., 1994a).” (VAM-SPC, 2008: 3)

Instances in which stakeholders proposedor submitted additional texts (notreferenced in the draft SAR) forconsideration and inclusion in the finalSAR.

Recommendation to includeadditional texts.

“The studies used in the screening assessment to estimatefetal or neonate plasma levels of free bisphenol A hadmethodological problems.” (BPA-SPC, 2008: 4)

Instances in which stakeholders criticizedtexts referenced in the draft SAR andproposed that they not be considered orincluded in the final SAR.

Recommendation to excludetexts.

“Releases of bisphenol A to the environment have beenoverstated. The assessment does not correctly representindustrial processes and makes unvalidatedassumptions.” (BPA-SPC, 2008: 9)

Instances in which stakeholders criticizedthe way in which texts referenced inthe draft SAR had been interpreted.

Criticism of aninterpretation of texts.

Government Response“The Government of Canada acknowledges the receipt of

information regarding additional scientific studies onbisphenol A. As screening assessments do not representexhaustive or critical reviews of all available data, onlythose studies deemed to influence the ultimateconclusions of the assessment have been added to theassessment report.” (BPA-SPC, 2008: 11)

Instances in which a proposed/submittedadditional text was not included.

Rebuffing arecommendation toinclude additional texts.

“The key studies . . . were rigorously designed, relevantand considered of high utility by the expert panel;therefore, they were included in hazard identificationand risk characterization.” (BPA-SPC, 2008: 15–16)

Instances in which a recommendation toexclude a referenced text was notaccepted.

Rebuffing arecommendation toexclude texts.

“We acknowledge that there are differences between theEU draft RAR (EU, 2007) and the Canadian draftscreening assessment. . . . With respect to exposure, thedifferences are greater in the final screening assessmentbecause data from a contemporary consumer productssurvey were used for determining Canadian exposures.”(VAM-SPC, 2008: 2)

Instances in which a criticism of aninterpretation was rejected.

Rebuffing a criticism of aninterpretation of texts.

“Modeling of exposures was modified based on the newresidue data provided.” (VAM-SPC, 2008: 7)

Instances in which a proposed/submittedadditional text was included.

Accommodating arecommendation toinclude additional texts.

“However, as an evaluation by SCOEL (2005) discountedthe Czajkowska et al. study due to poor documentation,and since the study has not been cited in more recentassessments of vinyl acetate (EU RAR, 2008 (EU, 2008),US EPA, 2006b), this study is no longer considered infinal screening assessment.” (VAM-SPC, 2008: 3)

Instances in which a recommendation toexclude a text was accepted.

Accommodating arecommendation toexclude texts.

“The Government was in agreement with this comment[Reliance on Section 71 submissions does notaccurately portray residues of vinyl acetate monomer inCanadian consumer products] and solicited industry totest consumer products in the North Americanmarketplace for residues of vinyl acetate monomer.”(VAM-SPC, 2008: 8)

Instances in which a criticism of aninterpretation was acknowledged.

Accommodating a criticismof an interpretation oftexts.


cepted knowledge, the continuity of organizationalactivities, and the use of codified norms as a basisfor action. The specific practices, identified fromtexts describing how risks are assessed and man-aged in Canada, included referencing an acceptedbody of scientific or regulatory knowledge, anchor-ing actions in organizational precedents or taken-for-granted routines, and categorizing chemicals inways that subsequently led to a predetermined se-quence of clearly delineated actions.

First, referencing refers to instances in which thesystematic application of accepted scientificknowledge and techniques is emphasized with ac-knowledgements of extant research, scientific ex-perts, and the work of other jurisdictions. For ex-ample, the Government website on chemicalsubstances noted: “The law says that chemical sub-stances must be examined in a scientific and thor-ough way” (Government of Canada, 2007b). A “sys-tematic,” “rigorous,” and “comprehensive”application of science was necessary “to make surethat every substance that could potentially affecthuman health or our environment” was identifiedfor further attention (Government of Canada,2007c). Descriptions of risk assessment and man-agement processes also emphasized the importanceof carrying out activities in ways that were consis-tent with those of international peers. Accordingly,the Government was “mindful of internationalstandards” (Environment Canada, 1995) and re-viewed those decisions of other countries that were“based on scientific considerations” (CEPA, 1999:45). The incorporation of extant scientific findingswas, therefore, a key part of risk assessment andmanagement insofar as descriptions emphasizedextensive referencing to published research andconclusions reached in other jurisdictions.

Second, anchoring refers to instances in whichrisk assessment and management processes wereexplicitly linked to past activities, decisions, andprecedents, thus emphasizing continuity. For ex-ample, descriptions stressed that processes derivedfrom longstanding laws, including CEPA (Govern-ment of Canada, 2007a), and that the current initia-tive was informed by prior risk assessmentactivities:

The Government of Canada has been doing risk as-sessment and management for many years. . . .There are some 60 different tools in place. (Govern-ment of Canada, 2010d)

In texts describing the Challenge, risk assessmentand management processes were thus “situated”among and represented as a continuation of priorroutines and experience.

Third, categorizing—assigning chemicals to pre-existing categories with clear rules of inclusion andfrom which certain actions followed—was also anintegral part of the risk assessment and manage-ment processes described in Government texts.“Categorization” was the official term used to de-scribe the initial screening of the 23,000 substances(Government of Canada, 2007b). If a chemical wascategorized as persistent, bioaccumulative, and in-herently toxic, further evaluation was then re-quired. If a chemical was then categorized as“CEPA-toxic,” particular actions were mandated.Thus, categorizing was central to risk assessmentand management processes because it established aseries of “if/then” protocols: if a chemical met thespecific rules of inclusion to be categorized in aparticular way, then clearly delineated actionsfollowed.

Finally, different forms of temporal sequencingwere also described in texts about the Challenge.The various steps of the formal process were asso-ciated with strict timelines and often representedin a linear, flowchart-like way. For example, CEPAstated that the Government “shall, within sevenyears from the giving of Royal Assent to this Act,categorize the substances that are on the DomesticSubstances List” (1999: 42). In 2005, the Govern-ment was “on track to meet its legislated deadlineof September 2006 for completing this categoriza-tion exercise” (Government of Canada, 2005). In theChallenge, each batch had specific timelines pre-sented in a range of visual formats. A web pageentitled “Deadlines associated with the Challenge”specified the “key milestones for each releasedBatch” (Government of Canada, 2010e).

In sum, normalizing describes a bundle of inter-related practices that collectively emphasize themindful application of accepted knowledge, thecontinuity of organizational activities, and the useof codified norms as a basis for action. An acceptedbody of scientific knowledge and precedents fromother jurisdictions or prior organizing activities in-formed risk assessment and management pro-cesses, which meant that taken-for-granted catego-ries were invoked and predetermined sequences ofactivities followed.

Problematizing. Our analysis also identified asecond interrelated set of practices, which we referto as problematizing, that was in tension with nor-malizing. Problematizing emphasized the reflexiveacknowledgement of potential inadequacies inknowledge, discontinuity in organizational activi-ties, and the use of open-ended deliberations as abasis for action. In contrast to the clear “roadmap”(Government of Canada, 2007a) provided by theapplication of practices associated with normal sci-


ence and precedent, descriptions of the Challengeunderlined that assessing and managing risks couldalso involve other practices. These practices in-cluded recognizing particular situations as uniqueor novel and innovating accordingly in conjunctionwith questioning the adequacy of existing scientificknowledge and acknowledging the plurality ofmultiple stakeholders involved in and affected byrisk assessment and management processes.

First, we found evidence of particularizing, in-stances of mentions of the need to single out indi-vidual chemicals as unique despite the systematicapplication of an accepted body of knowledge anduse of generalized categories. Texts describing theChallenge noted that “risk depends on what thechemical substance is, the quantity required tocause effects, the amount and length of time ofexposure, and how that exposure takes place (infood or air or water, for example)” (Government ofCanada, 2007d), suggesting idiosyncratic analysesfor each individual chemical.

Second, texts about the Challenge stressed thatrisk assessment and management also involved in-novating and hence the possibility of clear breakswith prior activities of the Government and otherjurisdictions. Such descriptions stood in contrast tothose emphasizing the anchoring of current pro-cesses in routine activities and stressing continuity,experience and familiarity. Canada’s risk assess-ment and management processes were not simplyan extrapolation of the past or an imitation of othercountries—they were described as novel anddifferent.

Canada, like the United States and European coun-tries, has been evaluating and managing chemicalsubstances for decades. However, Canada is the firstcountry in the world to categorize the thousands ofchemical substances in use before comprehensiveenvironmental protection laws were created. (Gov-ernment of Canada, 2007a)

Thus, claims of innovating were often juxtaposedand in tension with anchoring and referencing.

Third, we noted extensive questioning in textsdescribing the risk assessment and managementprocesses. Sometimes, questions appeared as a rhe-torical device to impose order; for example, manyof the questions posed on the website were directlyanswered, thus helping to construct certainty.However, we found numerous other instances inwhich questions were asked but not answered,thereby helping to construct uncertainty.

How and where are chemical substances getting intoour air, water and food, and at what levels are theyfound? How much exposure might we have to agiven chemical substance? What happens after its

use and disposal? What might short or long termexposure mean? What do advancements in researchtell us we did not know before? These and otherquestions guide Government of Canada scientists inresearching and assessing chemical substances.(Government of Canada, 2007a)

There were, consequently, many instances of rec-ognizing that uncertainty existed and informationwas incomplete or contradictory. Actions had to betaken “in a precautionary manner i.e., decisionmaking will not wait for scientific certainty in allinstances” (Government of Canada, 2007e) and “theabsence of information from industry and stake-holders will not preclude measures being taken tosafeguard human health and the environment”(Government of Canada, 2009d). Thus we identi-fied tensions between open-ended questioning,with its foregrounding of uncertainty, and the ref-erencing of extant scientific knowledge and un-problematic categorizing of chemicals based onthat knowledge.

Finally, we noted tensions between referencing,anchoring, and categorizing, which imply a processdominated by scientists, and pluralizing, which ac-knowledged that other stakeholders also had im-portant roles. For example, the Government bro-chure on assessing chemicals (Government ofCanada, 2005) named industry and the environ-mental community as important participants inrisk assessment and management processes. Otherdocuments emphasized plurality by using the ter-minology of “working together” and “partner-ships.” A major objective of the Challenge was“transparency, to involve all stakeholders, includ-ing members of the public in the process” (Keller &Heckman, 2007: 2–3).

In sum, problematizing describes a bundle of in-terrelated practices that include questioning theadequacy of a body of knowledge as well as draw-ing attention to unique circumstances requiring in-novation and consideration of plural perspectivesin open-ended deliberations. In emphasizing abreak from prior risk assessment and managementprocesses, it is in tension with normalizing.

Constructing the Meaning of Specific Chemicals

We examined our data on VAM and BPA forevidence of normalizing and problematizing prac-tices and found differences in the prominence ofthese practices in the construction of meaning ofthe two chemicals. The meaning of VAM emergedprimarily from practices associated with normaliz-ing. BPA texts also showed evidence of normaliz-ing practices, but, in addition, showed far greater


evidence of problematizing practices than didVAM texts.

VAM. Our analysis indicated that VAM was ini-tially particularized, by being “identified as a highpriority” (VAM-DSAR, 2008: 1). Following this,traces of normalizing practices became much moreapparent. For example, the conclusion of the draftassessment was explicitly anchored in a longstand-ing policy of the Government that was reaffirmedon the advice of the Challenge Advisory Panel inrelation to the first batch of chemicals consideredunder the Challenge. This policy established a gen-eral precedent: chemicals categorized as non-threshold carcinogens were automatically deemedto be toxic to humans because any level of exposurecan result in harm to human health. Hence, thecategorizing of VAM as having a nonthresholdmode of action led directly to the conclusion that itwas toxic to humans.

For substances for which the critical health effect isassumed to have no threshold of exposure for induc-tion, e.g., a genotoxic carcinogen, it is assumed thatthere is a probability of harm to human health at anylevel of exposure. (VAM-RMS, 2008: 1)

This document (on page 2) did acknowledge theexistence of uncertainty regarding a “full evalua-tion of the mode of action” but, otherwise, therewas relatively little mention of uncertainty or pre-caution. The draft SAR mentioned uncertainty onlysix times and precaution only twice, placing fargreater emphasis on weight of evidence.

The texts related to VAM contained extensivereferencing to other scientific texts. For example,the draft SAR referenced an unpublished draft riskassessment report produced by the EuropeanUnion (EU, 2007). However, it was concluded thatthe unofficial and “uncertain status” (Governmentof Canada, 2008a) of this text meant it was insuffi-cient to overturn the assumption of a nonthresholdmode of action. Submissions to the public consul-tation subsequently referenced a final version ofthis text published in 2008 (EU, 2008), as well asadditional scientific studies. The Government ac-knowledged in the SPC that new information hadcome to light regarding VAM’s mode of action,including the final, published version of this EUtext, which was referenced more than 50 times inthe final SAR.

Prior to issuing the final SAR, the Governmentappeared to particularize VAM once again when itasked the Panel whether it agreed that the weight ofevidence and application of precaution now sup-ported a conclusion that VAM was not toxic. ThePanel agreed that it did. In the final SAR, VAM wasrecategorized as having a threshold mode of action

and, since studies indicated that anticipated expo-sure levels did not approach the threshold, VAMwas not considered to be toxic. Accordingly, therewas no requirement for the Government to preparerisk management measures and, as far as the Chal-lenge was concerned, the file was closed. Table 4summarizes our findings for which practices weremost prominent in the construction of meaning forVAM, over time.

BPA. Our analysis indicated that BPA was alsoinitially particularized as “a high priority for as-sessment with respect to risks to human health”(BPA-DSAR, 2008: 4). Subsequent texts addressingBPA provide evidence that particularizing contin-ued insofar as BPA, as of August 2010, was the onlychemical in the Challenge to warrant its own linkfrom the Government’s chemicals managementportal (Government of Canada, 2010f) and its ownfact sheet and frequently asked questions pages(Government of Canada, 2009e).

There was also significant evidence of question-ing. For example, the draft SAR for BPA includesthe terms “uncertainty” or “uncertainties” 28 timesin 106 pages (in the draft SAR for VAM, uncertaintywas mentioned only 7 times in 44 pages). In thesections on effects on human health, where bothchemicals were found be toxic, these terms wereused 21 times in 39 pages for BPA as opposed to 5times in 4 pages for VAM. Similar patterns werenoted in the SPC: 17 times in 26 pages for BPA andonly once in 9 pages for VAM. In the final SAR,uncertainty was mentioned 28 times in 107 pagesin the case of BPA and only 4 times in 47 pages inthe case of VAM. Not only was uncertainty men-tioned more often, but also greater attention wasdrawn to it. For example, in the draft SAR for BPA,the heading “Uncertainties in Evaluation of Risk toHuman Health and Identification of ResearchNeeds” (BPA-DSAR, 2008: 72) was followed by thestatement, “There are a number of uncertaintiesrelated to evaluation of risk to human health,”which, in turn, was followed by seven bulletpoints, in each of which uncertainty/uncertaintiesappeared at least once. This structure was retainedin the final SAR. In other words, uncertainty as toBPA’s toxicity was foregrounded. (The final VAMdocument had only one heading mentioning uncer-tainties, which was followed by two paragraphscontaining no bullet points and only one instanceof “uncertainties.”)

In the case of BPA, uncertainty was consistentlylinked to precaution, which was also mentionedmore often than with VAM: 10 times in the draftSAR, 15 times in the SPC, and 9 times in the finalSAR (compared to 2, 4, and 1 for VAM). The SPC onBPA stated:


TA

BL

E4

Pra

ctic

esM

ost

Pro

min

ent

inth

eC

onst

ruct

ion

ofM

ean

ing

for

VA

Mov

erT

ime

Dra

ftS

AR

(DS

AR

)M

ay20

08R

MS

May

2008

CA

PR

epor

tO

ctob

er20

08S

PC

Nov

embe

r20

08F

inal

SA

R(F

SA

R)

Nov

embe

r20

08W

ebS

um

mar

y

-“B

ased

pri

nci

pal

lyon

the

wei

ght

ofev

iden

ceev

alu

atio

nof

IAR

C(1

995)

,”cr

itic

alef

fect

isca

rcin

ogen

icit

y.-

Foc

us

onn

onth

resh

old

mod

eof

acti

on,

wh

ich

isco

nst

ruct

edas

the

app

rop

riat

eca

tego

riza

tion

base

don

avai

labl

ein

form

atio

n.

-L

ittl

em

enti

onof

un

cert

ain

tyor

pre

cau

tion

.

-S

ome

men

tion

ofri

skm

anag

emen

tm

easu

res

tobe

take

nif

VA

Mis

fou

nd

top

ose

risk

s,bu

tfe

wd

etai

lsgi

ven

.-

Foc

us

onn

onth

resh

old

mod

eof

acti

onal

thou

ghin

com

ple

tein

form

atio

nac

know

led

ged

.

-P

anel

isas

ked

spec

ific

ques

tion

inre

lati

onto

VA

M:

Doe

sth

eP

anel

agre

ew

ith

the

app

roac

hto

app

lica

tion

ofw

eigh

tof

evid

ence

and

pre

cau

tion

?

-G

over

nm

ent

invo

kes

text

sby

outs

ide

acto

rs,

esp

ecia

lly

pu

blis

hed

risk

asse

ssm

ent

rep

ort

(EU

,20

08).

-G

over

nm

ent

ackn

owle

dge

sp

revi

ous

“in

com

ple

ten

ess”

ofin

form

atio

nan

dre

fere

nce

s“n

ewd

ata.

”-

Lit

tle

men

tion

ofu

nce

rtai

nty

.

-R

ever

ses

dra

ftS

AR

con

clu

sion

.-

“Bas

edp

rin

cip

ally

onth

ew

eigh

tof

evid

ence

eval

uat

ion

ofIA

RC

(199

5)an

dta

kin

gin

toco

nsi

der

atio

nm

ore

rece

nt

dat

ain

clu

din

g(E

U,

2008

),”

crit

ical

effe

ctis

carc

inog

enic

ity.

-F

ocu

son

thre

shol

dm

ode

ofac

tion

,w

hic

his

con

stru

cted

asth

eap

pro

pri

ate

cate

gori

zati

onba

sed

onn

ewin

form

atio

n.

-V

AM

“is

not

con

sid

ered

tobe

har

mfu

lto

hu

man

hea

lth

.”-

“Th

isd

ecis

ion

was

base

don

new

info

rmat

ion

rece

ived

du

rin

gth

ep

ubl

icco

mm

ent

per

iod

,as

wel

las

mor

ere

cen

tin

form

atio

nfr

omth

eri

skas

sess

men

tco

nd

uct

edby

the

Eu

rop

ean

Un

ion

.”

Pro

blem

atiz

ing

Par

ticu

lari

zin

g(V

AM

isa

pri

orit

ym

erit

ing

its

own

asse

ssm

ent)

.S

ome

ques

tion

ing

(in

form

atio

nab

out

mod

eof

acti

onin

com

ple

te).

Nor

mal

izin

gR

efer

enci

ng

(wei

ght

ofev

iden

ceof

IAR

C;

cita

tion

ofE

U(2

007)

risk

asse

ssm

ent

rep

ort)

.A

nch

orin

gan

dca

tego

rizi

ng

(as

alo

ngs

tan

din

gp

olic

y,an

din

lin

ew

ith

pan

elad

vice

for

batc

h1,

non

thre

shol

dca

rcin

ogen

sar

eca

tego

rize

das

toxi

c).

Nor

mal

izin

gR

efer

enci

ng

(wei

ght

ofev

iden

ceas

sess

men

tof

IAR

C);

An

chor

ing

and

cate

gori

zin

g:(n

onth

resh

old

mod

eof

acti

onre

sult

sin

toxi

city

con

clu

sion

)S

ome

pro

blem

atiz

ing

Qu

esti

onin

g(n

onth

resh

old

mod

eof

acti

on“c

ann

otbe

pre

clu

ded

”).

Pro

blem

atiz

ing

Par

ticu

lari

zin

g(V

AM

sin

gled

out

ina

spec

ific

ques

tion

rela

tin

gto

its

asse

ssm

ent

only

,n

ota

mor

ege

ner

alp

olic

yis

sue)

Nor

mal

izin

gE

xten

sive

refe

ren

cin

g(i

nvo

kin

gn

ewsc

ien

tifi

cte

xts)

.E

xten

sive

anch

orin

gan

dca

tego

rizi

ng

(dra

ftfi

nd

ing

was

app

rop

riat

egi

ven

lon

gsta

nd

ing

pol

icy

and

earl

ier

Pan

elad

vice

re.

chem

ical

sca

tego

rize

das

non

thre

shol

dca

rcin

ogen

s;re

vers

alof

dra

ftfi

nd

ing

isal

soap

pro

pri

ate

owin

gto

reca

tego

riza

tion

).

Nor

mal

izin

gE

xten

sive

refe

ren

cin

g(n

ewd

ata

now

avai

labl

e).

Cat

egor

izin

g(a

dif

fere

nt

mod

eof

acti

onas

com

par

edto

dra

ftS

AR

);an

chor

ing

(in

acce

pte

dp

ract

ice

for

asse

ssin

gth

resh

old

carc

inog

ens,

i.e.

,vi

aex

pos

ure

scen

ario

s).

Nor

mal

izin

gE

xten

sive

refe

ren

cin

g(n

ewin

form

atio

nre

ceiv

edan

dE

Uas

sess

men

t).

It is important to note that precaution may be in-voked when there is uncertainty about the extent towhich the available evidence actually indicates thatthe substance is causing such harm. (BPA-SPC,2008: 17).

It was, therefore, “considered appropriate to applya precautionary approach when characterizingrisk” (BPA-DSAR, 2008: 2–3). Uncertainty aboutBPA’s toxicity to both humans and the environ-ment “warranted” the use of a precautionary ap-proach in evaluating risk (BPA-FSAR, 2008:32–33).

Precaution was foregrounded in a number ofways. First, the Panel was consulted prior to issu-ing the draft SAR on whether the weight of evi-dence and application of precaution supported theGovernment’s conclusion of toxicity, and the Panelagreed. Second, a series of authoritative texts, au-thored by the Government and supporting its inter-pretation of precaution, were referenced in theSPC. Third, the Government emphatically invokedits right—and duty—to exercise judgment: it was“incumbent on the government to judge whether asubstance has the potential to cause serious or ir-reversible damage to the environment and/or hu-man health (BPA-SPC, 2008: 18). Even when theGovernment concluded that exposure levels fornewborns and infants were “below those that couldcause health effects,” it stated that it would still act“to further limit exposure” because of the “uncer-tainty raised in some studies” (Government of Can-ada, 2009f). In other words, BPA was portrayed as avery particular case insomuch as the Governmenthad an obligation to act independently of otherregulatory agencies and to take the lead in restrict-ing BPA. In doing so, it engaged in pluralizing byrecognizing that the public had an important stakein this outcome, as did other stakeholders, mostnotably industry (Government of Canada, 2008).For example, the Government stated it would “sup-port manufacturers in the evaluation of replace-ment options” (BPA-RMS, 2008: 7) and “continueto work with the food packaging industry and in-fant formula manufacturers” (Health Can-ada, 2008a).

Fourth, precaution was linked to actions (i.e.,risk management measures) required to deal withBPA’s threat to human health and the environment.The eight-page RMS also provided an extensivediscussion of risk management measures to be con-sidered in the event that BPA was found to poserisks—six paragraphs amounting to nearly a page oftext (compared to a single paragraph in the compa-rable six-page VAM document, even though thedraft SAR found VAM to be toxic to human health).

In linking precaution to action, texts indicated thatspecific subpopulations of Canadians were at risk:“the pregnant woman/fetus and infant” (BPA-RMS,2008: 3). This distinction was further emphasizedin the SPC, which referred to “vulnerable popula-tions” (BPA-SPC, 2008: 2) while underlining thatmost products containing BPA posed “little risk toCanadians” (BPA-SPC, 2008: 3). A press releasestressed that “the general public need not be con-cerned” (Health Canada, 2008b). However, at-tempts to construct boundaries between subpopu-lations at risk and those not at risk appear to havebeen problematic. The web-based summary (Gov-ernment of Canada, 2008c) stated that BPA “did notpose a risk to the general population” but then wenton to list a series of “precautionary measures” thatall Canadians could take to avoid risk.

BPA’s meaning—as toxic—resulted from greateruse of problematizing practices than in the case ofVAM. In addition to containing traces of the prac-tices associated with normalizing, BPA texts alsoconstructed BPA as an exceptional case: uncer-tainty regarding its toxicity for a particularly vul-nerable subpopulation warranted a precautionaryapproach and hence actions to manage risks; andthe Government asserted its right to take the lead inmaking innovative, independent judgments aboutBPA where necessary. Table 5 summarizes ourfindings for which practices were most prominentin the construction of meaning for BPA, over time.

Discursive Work: Stabilizing, Destabilizing, andRestabilizing Meanings

We examined how the various meanings of VAMand BPA were held in place. The meanings of bothchemicals were initially stabilized in the draftSARs as toxic. In each case, some stakeholdersattempted to destabilize this meaning with a viewto subsequently restabilizing the chemical’s mean-ing as safe. In the case of VAM, destabilizing at-tempts were successful, and VAM’s meaning astoxic was destabilized, then changed and subse-quently restabilized as safe. In the case of BPA,destabilizing attempts failed, and BPA’s initialmeaning as toxic remained stabilized. In this sec-tion, we show how the amount and nature of dis-cursive work carried out by stakeholders and theGovernment differed according to the prevalence ofnormalizing or problematizing practices.

VAM. The initial assessment of VAM, which wasassociated primarily with normalizing practices,served to stabilize its meaning—momentarily atleast—as toxic; VAM was categorized as a non-threshold carcinogen that, according to precedent,led automatically to a conclusion of toxicity.


TA

BL

E5

Pra

ctic

esM

ost

Pro

min

ent

inth

eC

onst

ruct

ion

ofM

ean

ing

for

BP

Aov

erT

ime

CA

PR

epor

tM

arch

2008

Dra

ftS

AR

(DS

AR

)A

pri

l20

08R

MS

Ap

ril

2008

SP

CO

ctob

er20

08F

inal

SA

R(F

SA

R)

Oct

ober

2008

Web

Su

mm

ary

-P

anel

isas

ked

spec

ific

ques

tion

inre

lati

onto

BP

A:

Doe

sth

eP

anel

agre

ew

ith

the

app

roac

hto

app

lica

tion

ofw

eigh

tof

evid

ence

and

pre

cau

tion

?

-C

riti

cal

effe

ctis

“rep

rod

uct

ive

and

dev

elop

men

tal

toxi

city

”an

dd

ata

set

“th

ough

hig

hly

un

cert

ain

,is

sugg

esti

veof

pot

enti

alef

fect

sat

dos

es”

sim

ilar

toex

pos

ure

s.-

Un

cert

ain

tyis

pro

min

ent

and

lin

ked

top

reca

uti

on,

wh

ich

isco

nst

ruct

edas

the

app

rop

riat

ere

spon

seso

that

acti

onca

nbe

take

nin

rela

tion

tose

nsi

tive

subp

opu

lati

ons,

i.e.

,“t

he

pre

gnan

tw

oman

/fet

us

and

infa

nt.

”

-E

xpli

cit

and

rep

eate

dm

enti

onof

pre

cau

tion

ary

app

roac

h,

wh

ich

isli

nke

dto

acti

onin

rela

tion

to“s

ensi

tive

subp

opu

lati

ons”

-A

ctio

nou

tlin

edin

quit

ed

etai

led

man

ner

:“s

up

por

tm

anu

fact

ure

rsin

the

eval

uat

ion

ofre

pla

cem

ent

opti

ons”

for

infa

nt

form

ula

food

pac

kagi

ng;

and

ban

baby

bott

les.

-G

over

nm

ent

invo

kes

its

own

text

sin

sup

por

tof

pre

cau

tion

ary

acti

on;

and

its

righ

tto

mak

eju

dgm

ents

inli

ght

ofu

nce

rtai

nty

and

vuln

erab

lesu

bpop

ula

tion

s-

Gov

ern

men

t“i

sle

adin

gth

ed

evel

opm

ent

ofri

skas

sess

men

tm

eth

odol

ogy

for

succ

inct

and

focu

sed

asse

ssm

ents

tobe

tter

add

ress

pu

blic

and

scie

nti

fic

con

cern

sin

mor

eef

fici

ent

man

ner

.”

-A

ffir

ms

dra

ftS

AR

con

clu

sion

-C

riti

cal

effe

ctis

“rep

rod

uct

ive

and

dev

elop

men

tal

toxi

city

”;an

dd

atas

et“t

hou

ghh

igh

lyu

nce

rtai

n,

issu

gges

tive

ofp

oten

tial

effe

cts

atd

oses

”si

mil

arto

exp

osu

res.

-U

nce

rtai

nty

isp

rom

inen

tan

dli

nke

dto

pre

cau

tion

,w

hic

his

con

stru

cted

asth

eap

pro

pri

ate

resp

onse

soth

atac

tion

can

beta

ken

inre

lati

onto

sen

siti

vesu

bpop

ula

tion

s,i.

e.,

“th

ep

regn

ant

wom

an/f

etu

san

din

fan

t.”

-B

PA

isto

xic.

-“C

hil

dre

n,

incl

ud

ing

new

born

san

din

fan

ts”

are

atri

sk.

-B

PA

“doe

sn

otp

ose

ari

skto

the

gen

eral

pop

ula

tion

,”h

owev

er“i

fC

anad

ian

sh

ave

con

cern

s”a

seri

esof

pre

cau

tion

ary

mea

sure

sis

pro

vid

ed.

-N

ewri

skm

anag

emen

tm

easu

res

are

tobe

take

n.

-G

over

nm

ent

wor

kin

gw

ith

ind

ust

ry.

Pro

blem

atiz

ing

Par

ticu

lari

zin

g(B

PA

sin

gled

out

ina

spec

ific

ques

tion

rela

tin

gto

its

asse

ssm

ent

only

,n

ota

mor

ege

ner

alp

olic

yis

sue)

.

Pro

blem

atiz

ing

Par

ticu

lari

zin

g(B

PA

isp

rior

ity

mer

itin

git

sow

nas

sess

men

t;vu

lner

able

subp

opu

lati

ons

affe

cted

).E

xten

sive

ques

tion

ing

(un

cert

ain

ty)

Pro

blem

atiz

ing

Ext

ensi

vequ

esti

onin

g(u

nce

rtai

nty

).In

nov

atin

g(n

ewri

skm

anag

emen

tac

tion

s).

Plu

rali

zin

g(w

orki

ng

wit

hin

du

stry

).

Pro

blem

atiz

ing

Ext

ensi

vequ

esti

onin

g(u

nce

rtai

nty

),in

nov

atin

g(g

over

nm

ent

isth

ele

ader

inas

sess

men

tm

eth

ods

and

pre

cau

tion

ary

acti

on).

Par

ticu

lari

zin

g(v

uln

erab

lesu

bpop

ula

tion

saf

fect

ed).

Som

en

orm

aliz

ing:

anch

orin

g(r

efer

enci

ng

ofG

over

nm

ent

text

son

pre

cau

tion

).

Pro

blem

atiz

ing

Par

ticu

lari

zin

g(v

uln

erab

lesu

bpop

ula

tion

saf

fect

ed).

Ext

ensi

vequ

esti

onin

g(u

nce

rtai

nty

).

Pro

blem

atiz

ing

Inn

ovat

ing

(new

risk

man

agem

ent

mea

sure

s).

Plu

rali

zin

g(w

orki

ng

wit

hin

du

stry

;ac

know

led

gin

gco

nce

rns

ofC

anad

ian

s).

Som

en

orm

aliz

ing

Cat

egor

izin

g(b

oun

dar

ies

nar

row

lyd

raw

nar

oun

dri

skbe

arer

s,i.

e.,

infa

nts

,n

otge

ner

alp

opu

lati

on).

Stakeholders responded to that conclusion by at-tempting to destabilize it and restabilize the mean-ing of VAM as safe. They did so by submitting textsrecommending that additional texts should be ref-erenced in the final SAR, most notably the 2008version of the EU’s final risk assessment report (EU,2008), which proposed that VAM had a thresholdmode of action.

[Submitters] argued that the draft screening assess-ment conclusion had been overly precautionary anddid not consider the weight of evidence of newtoxicology data. In particular, the submitters recom-mended that Health Canada consider a thresholdmode of action for the carcinogenicity of vinyl ace-tate which has been recently proposed in the scien-tific literature and in risk assessments from Europe[i.e., EU, 2008]. (VAM-SPC, 2008: 1)

The Government accommodated this criticism andreferenced this text in the final SAR, concludingthat there was a threshold mode of action, even ifVAM was one of few “exceptional cases wheregenotoxic action is thought to be thresholded”(SCHER, 2008: 5). Accordingly, VAM was allocatedto a different category—it was now considered to bea threshold carcinogen. The original sequencingthat followed from the nonthreshold categorizationno longer applied, and VAM could no longer auto-matically be categorized as toxic. This change ofcategorization also meant that, in accordance withnormal scientific practice and prior assessments,additional information was required to ascertainwhether the exposure of Canadians was likely to beabove or below the threshold at which harm wascaused. Critics therefore recommended the refer-encing of additional texts that addressed exposure.The Government also accommodated these recom-mendations, referencing 21 additional texts in thefinal SAR, and concluding that exposure to VAMwas expected to be low and was not considered tobe harmful to human health.

Our analysis illustrates how normalizing prac-tices structured attempts to destabilize VAM’smeaning, which was dictated primarily by whetheror not it was categorized as a threshold or non-threshold carcinogen, as well as the Government’sresponse. Normalizing established a relatively cleartarget for those wishing to destabilize VAM’s mean-ing as toxic: adding scientific texts that could leadto recategorization. Once recategorized, the newmeaning of VAM then emerged from the new se-quencing that followed from this new categoriza-tion, methods anchored in prior assessments, andreferencing additional texts that developed expo-sure scenarios. So, although Government often re-buffed criticisms, those instances in which it ac-

commodated them (by referencing additional texts)led to the destabilizing of VAM’s meaning as toxicand the restabilization of a new meaning—VAMas safe.

BPA. The initial assessment of BPA was associ-ated with normalizing practices, but it also showedfar greater evidence of problematizing practicesthan did the draft SAR for VAM. These practicesserved to stabilize BPA’s meaning as toxic not be-cause it belonged to a specific category, but becauseits effects were uncertain, and this uncertainty waslinked, through the precautionary principle, to theneed for action to manage potential risks. Attemptsto destabilize the meaning of BPA as toxic involvedprimarily challenges to the Government’s interpre-tation of certain texts and demands to exclude cer-tain texts that had been cited in the draft SAR.

First, stakeholders refuted the Government’s in-terpretation of texts describing the precautionaryprinciple and equating it with the need for action:

The application of the precautionary principle isexcessive and without an appropriate scientific ba-sis. Application of the weight-of-the evidence ap-proach and precautionary principle does not meanthat any uncertainty requires action as a precaution.The precautionary principle should be appliedwhen the weight of the evidence suggests that apotential threat to the environment and humanhealth exists and when that threat is of serious orirreversible damage. Until both of these conditionsare met, application of the precautionary principleto justify actions limiting trade is inappropriate.(BPA-SPC, 2008: 18)

The Government rebuffed this criticism by invok-ing a series of extant texts that it had previouslyauthored to argue that its interpretation—involvingthe need for action—was entirely appropriate in thelight of ongoing questioning and uncertainty re-garding BPA’s risk to human health:

A precautionary approach to decision making, asdefined in A Framework for the Application of Pre-caution in Science-Based Decision Making AboutRisk (Government of Canada), emphasizes the needto take appropriate action, even in the absence offull scientific demonstration of cause and effect.(BPA-SPC, 2008: 18)

It also asserted its right to act in the face of scien-tific uncertainty and linked this to innovating.

The Government of Canada is leading the develop-ment of risk assessment methodology. . . . For sub-stances where no other regulatory bodies have con-ducted a rigorous assessment, the Government ofCanada will develop methodology to efficiently ad-dress the issues surrounding the substance (forexample, developing models to determine the expo-


sure of Canadians to a particular substance). (BPA-SPC, 2008: 14–15)

The Government thus justified acting differentlyfrom other jurisdictions because it was a leader,because of questions surrounding BPA’s health ef-fects, and because of the need to adopt a precau-tionary approach in protecting Canadians.

Additional criticisms that the Government’s in-terpretations of texts were value-driven were alsorebuffed, not by denying that they were value-driven, but on grounds that they were—and oughtto be—value-driven. For example, a repeated accu-sation that the Government was “too conservative”in its interpretation of texts on exposure estimateswas met with responses that it was appropriate tobe conservative.

The screening assessment must adopt a conservativeapproach when representing the potential for riskfrom wastewater treatment plant effluents in Can-ada. (BPA-SPC, 2008: 10)

It is also the practice to utilize conservative ap-proaches to estimate exposures to ensure that theenvironment is protected. (BPA-SPC, 2008: 11)

Canadian studies presented in the screening assess-ment are considered appropriate to represent Cana-dian conditions and to enable conservative-level as-sessment. (BPA-SPC, 2008: 17)

Thus, the Government’s interpretations of particu-lar texts were defended.

Second, attempts to destabilize the meaning ofBPA as toxic also took the form of recommenda-tions to exclude texts referenced in the draft SARon grounds that they were not scientifically valid orappropriate. The Government rebuffed suchcriticism:

The limitations and inconsistencies associated withthe mentioned datasets were recognized and ac-knowledged in the draft screening assessment; how-ever, the key studies identified in the developmen-tal neurotoxicity dataset were conducted by theNational Toxicology Program–Center for the Evalu-ation of Risks to Human Reproduction (NTP-CERHR) expert panel (a science-based interagencyprogram in the United States). These studies wererigorously designed, relevant and considered ofhigh utility by the expert panel; therefore, they wereincluded in hazard identification and risk character-ization. Scientists within and external to the Gov-ernment of Canada having the required expertisewere consulted for peer review and had significantinput on the validation of the scientific evidenceincluded in the draft screening assessment. (BPA-SPC, 2008: 15)

In this case, the Government used normalizingpractices to defend its inclusion and referencing of

certain texts by anchoring its activities in acceptedscientific protocols and peer review and by refer-encing other jurisdictions.

Our analysis shows how problematizing prac-tices structured attempts to destabilize the meaningof BPA as toxic insofar as critics attacked the linkbetween uncertainty, precaution, and action byquestioning Government interpretations. This wasdifficult, however, because the Government coulduse value-based inferences to justify its selectionand interpretation of texts, as well as its right to act.The particular nature of BPA, the questions sur-rounding its effects, the inadequacy of existing re-search, and the need to consider a plurality of in-terests were all grounds for Government action. Inaddition, whereas additional texts could resolvethe uncertainty concerning the categorization ofVAM as a nonthreshold carcinogen, additionaltexts could only increase the uncertainty in thecase of BPA. Additional texts referring to new evi-dence could not “undo” the existing evidence thatindicated BPA might not be safe; they merelyadded to the controversy. So, even though 26 addi-tional texts were referenced in the final SAR, BPA’smeaning as toxic was not destabilized—uncer-tainty still remained, and action was still required.

DISCUSSION AND CONCLUSION

In combining a performative process approach(Tsoukas & Chia, 2002), in which much of the workto date has been theoretical and empirical studiesare rare (Van de Ven & Poole, 2005), with a discur-sive perspective, our study contributes important,empirically grounded insights into the relationshipbetween organizing and risk.

A Performative Process Perspective on Risk: TheRole of Practices and Discursive Work

Our research questions asked how organizingprocesses construct the meaning of risk objects and,within these processes, what is the role of discur-sive work in stabilizing (and destabilizing) mean-ings? Our study illustrates how two different formsof social ordering emerge in risk assessment andmanagement processes, depending on how prac-tices are collectively enacted and unfold over time,and, in so doing, bring the past to bear on thepresent in specific ways. These forms of social or-dering—normalizing and problematizing—struc-ture the discursive work of actors in their attemptsto stabilize meanings as well as to destabilize andchange them. Objects therefore “become” risky orsafe in different ways. Note that although our anal-ysis distinguishes these two forms of social order-


ing, we do not wish to imply that they are mutuallyexclusive: normalizing practices were found in re-lation to both VAM and BPA, but in the case of BPAproblematizing practices were also prominent. Nordo we wish to imply that problematizing necessar-ily leads to findings of toxicity: in using this term,we are not referring to the problematization of thestatus of a given chemical as safe but, rather, to theproblematization of the codified norms on whichsuch categorizations are usually made. Similarly,normalizing does not necessarily lead to findings ofsafety; nor does it simply refer to practices thathave come to be routine or habitual, but to specificpractices that include the application of normalscience.

When normalizing practices dominate, the mean-ing of an object is constructed through discursivework (producing, distributing, and consumingtexts), much of which has taken place in the past.Earlier discursive work creates shared understand-ings as to accepted “facts,” causal models, catego-ries and their consequences, as well as to methodsfor generating and validating knowledge. In thisway, the meaning of a material object—in our case,the meaning of VAM as both toxic and safe (but atdifferent points in time)—is held in place by anexus of preexisting texts, such as scientific arti-cles, policy documents, and risk assessments fromother jurisdictions, that reflect the outcomes of dis-cursive work carried out in the past. Referencingscientific research, anchoring in past organiza-tional practice, categorizing with accepted rules ofinclusion and exclusion, and sequencing accordingto predetermined steps reproduce the authority andapplicability of a body of risk knowledge and reaf-firm the credibility of scientific experts. In otherwords, the rules of the “the risk game” (Slovic,1998) are understood and shared.

The weight of the past that is brought to bearthrough normalizing practices means there is rela-tively little need for contemporaneous discursivework to explain or justify what is occurring, evenwhen meanings are destabilized and changed. Inour case, VAM’s meaning changed from safe totoxic then back to safe again. It was 1 of only 2chemicals in over 120 assessed as of August 2010that, following an initial conclusion of toxicity inthe draft assessment, subsequently became safe.Yet this reversal was not controversial—and re-quired little contemporaneous discursive work—since the destabilization of the initial meaning ofVAM and its restabilization as safe were bothachieved through normal science and organiza-tional precedent: an additional text produced inanother jurisdiction was referenced, and the subse-quent recategorizing of VAM as a threshold carcin-

ogen then led to a different set of consequencesinvolving the referencing of other additional textsaddressing human exposures. Normalizing thus notonly helps to stabilize meanings, but also estab-lishes a relatively clear target for discursive workrequired in attempts to destabilize andchange them.

Our approach draws attention to how apparentlystabilized meanings of risk objects are, in fact, pre-carious, even when supported by normalizing prac-tices. VAM’s restabilized meaning as safe, althoughit may appear to have been fixed as a result of theEU document, is not immutable. The ScientificCommittee on Health and Environmental Risks(SCHER), which provides the EU with independentscientific advice, noted concerns with VAM’s cat-egorization in the EU text.

[VAM] is a very special compound in that it isgenotoxic, induces tumours in 2 animal species atlocalizations relevant to man. This would requireclassification as a Category 2 carcinogen. . . . In-stead, the RAR [EU, 2008] proposes classification asa category 3 carcinogen . . . [which] is not in accor-dance with the criteria for classification of carcino-gens. (SCHER & CAS, 2008: 5)

Consequently, some organizations continue tomaintain that VAM is toxic. For example, the envi-ronmental NGO Reach for Unbleached! hasclaimed that VAM “is a carcinogen and genotoxinwith no known safe threshold [and] its use in foodpackaging is inappropriate and should be disal-lowed” (Reach for Unbleached! 2008). So furtherdiscursive work may be required to maintainVAM’s meaning as safe if other stakeholders con-tinue in their attempts to destabilize and change it.

Our study also illustrates how problematizingpractices can be used to stabilize the meaning ofrisk objects. In the case of BPA, in which normal-izing practices were evident but eclipsed by con-siderable problematizing, this chemical “became”risky and remained so, despite discursive struggleand even though no other country had, at that time,found BPA to pose significant risks. When prob-lematizing practices dominate, the past weighs lessheavily and, as a result, considerably more contem-poraneous discursive work is required. The author-ity and applicability of a body of risk knowledgeare called into question, thereby constructing a par-ticular problem requiring innovative action. In thecase of BPA, a great many texts were produced,distributed, and consumed to elaborate a link be-tween uncertainty, precaution, and action to avoidpotential risks. Compared to the case of VAM, inthe BPA case texts were longer, more numerous,and involved far more questioning of authors’ le-


gitimacy, basic “facts,” causal models, and meth-ods for generating and validating knowledge. Inother words, with problematizing, what constitutesrelevant knowledge, how to produce it, and what todo with it are up for grabs—the rules of the “riskgame” (Slovic, 1998) are more fluid.

Our study highlights how problematizing prac-tices are associated with actors engaging in morediscursive work and for longer periods of time.Instead of struggle occurring within a paradigm,where issues of epistemology, valid evidence, andappropriate value judgments are settled, as withnormalizing, problematizing involves struggle oc-curring over a paradigm. Problematizing thusobliges actors to engage in considerable discursivework, whether their intention is to stabilize anddefend or to destabilize and change meanings, be-cause of the many fronts that are opened up in thediscursive struggle. In the case of BPA, the Govern-ment continued to produce and distribute texts fol-lowing the release of the final SAR, including awebpage, press releases, and updates on the ongo-ing research on BPA’s effects. This is because prob-lematizing practices open up possibilities: if anobject is particularized, then it is always possible toafford it further special treatment; if innovating isconstructed as necessary, then it is always possibleto invent novel ways of dealing with risk; if ques-tioning challenges the applicability of a body ofknowledge, then it is always possible to expand thescope of this questioning; and if interests are plu-ralized, then it is always possible to find alternativeways to balance stakeholder concerns. This gener-ates notable discursive work as risk assessors andmanagers explain their actions and attempt to setbounds on problematizing.

Stabilizing Meanings of Risk Objects throughNormalizing and Problematizing

The coalescing of practices into particular formsof social ordering is not random insofar as practicesare interrelated in ways that are meaningful andfamiliar to actors engaged in organizing processesbecause these bundles emanate from the past. Inother words, only a delimited number of forms ofsocial ordering will “make sense” in a particularsetting. Our study indicates the importance of nor-malizing and problematizing in the context of riskassessment and management processes. That nor-malizing practices are used in stabilizing meaningis, perhaps, not surprising, given the way in whichrisk is conceived of in late modernity: realist ap-proaches dominate; scientific rationality is perva-sive, and the scientific method is taken-for-granted(Gephart et al., 2009; Malenfant, 2009; Miller, 2009;

Topal, 2009). That normalizing practices can beused to destabilize and change meanings is lessobvious. Our study highlights how meanings mustfirst be destabilized before they can be changed andrestabilized: objects are always “becoming” evenwhen they are constructed primarily through tak-en-for-granted scientific practices and texts.

Problematizing practices can also be harnessedin attempts to stabilize and destabilize meanings.Our study indicates how practices and discoursecombine to make this form of ordering relevant tocontemporary risk assessment and managementprocesses through the way in which the dominantdiscourse that emphasizes “sound science” hasbeen recently challenged by the discourse of pre-caution (Andrée, 2005). The latter creates different“conditions of possibility” (Maguire & Hardy, 2006:8). Specifically, it allows for the stabilizing ofmeaning because of and despite uncertainty. Be-fore the emergence of this discourse, uncertaintywas largely understood to arise from data unavail-ability that rendered impossible “the assignment ofprobabilities to a defined set of negative outcomesand thus the calculation of ‘risk,’ formally defined”(Maguire & Ellis, 2009: 121); and it was assumedthat uncertainty could be eliminated by gatheringmore data. The discourse of precaution advocates abroader understanding of uncertainty as “incerti-tude”—that is, “the general inability to make reli-able predictions of damages” (Klinke & Renn, 2001:161)—a concept that includes not only formal no-tions of uncertainty, but also ambiguity and igno-rance. In so doing, the discourse of precaution fills“the vacuum created by a science that continuallysearches for certainty but which continually fails todeliver” (Adams, 2002: 311) by providing a “stra-tegic resource” (Hardy, Palmer, & Phillips, 2000)for actors to use in their discursive work to stabilizemeanings and construct risk objects in situations ofambiguity.

In summary, our study’s findings illustrate how,in the context of contemporary risk assessment andmanagement processes, the meanings of risk ob-jects can be stabilized in (at least) two ways: byinvoking certainties that are known and accepted,including scientific findings, precedents, and rou-tines, through normalizing practices; or by invok-ing uncertainties and the need to manage risks in aprecautionary manner, through problematizingpractices.

Organizing Processes in Other Contexts

Normalizing and problematizing are forms of so-cial ordering that are highly relevant to the pro-spective assessment and management of chemical


risks, a highly mediated process in which “risk islocalized in the sphere of physical and chemicalformulas” (Beck, 1992: 21). Different forms of socialordering may apply when risks are assessed andmanaged through organizing processes with a dif-ferent temporal orientation, and our study suggestsinteresting areas for future study. One possibilityfor research concerns situations in which risk isassessed and managed in situ—without mediationand in real time—as, for example, in “high-reliabil-ity” organizations such as aircraft carriers, oil rigs,nuclear submarines, and others (e.g., Leveson, Du-lac, Marais, & Carroll, 2009; Sullivan-Taylor & Wil-son, 2009). Here, risks are often immediate andborne by the very same actors who are assessingand managing them. In such contexts, a systematic,top-down approach to risk is often advocated(Leveson et al., 2009), including continuous moni-toring of devices measuring physical parameters,regular inspecting and record keeping, auditingprevious failures, and authorizing activitiesthrough successive levels of hierarchy. Such prac-tices appear to bundle into a form of social orderingthat might be termed “controlling.” At the sametime, a very different form of social ordering, per-haps termed “sensing,” might also be importantinsomuch as actors rely on practices such as intu-iting, in which they use experience and tacitknowledge to process weak signals to conclude thatsomething is wrong (Weick & Sutcliffe, 2001)—forinstance, they actively look for unusual situationsand listen for strange sounds.

Risk is also assessed and managed retrospec-tively, such as in hearings and inquiries (e.g., To-pal, 2009; Winch & Maytorena, 2009). Here, organ-izing practices involve a forensic component todetermine causality, involving interviewing, cross-examining, and reconstructing event time lines.These practices could bundle into a form of socialordering that might be termed “investigating.” Ad-ditionally, because inquiries are often intended toreassure the public, they may also feature a form ofsocial ordering that could be termed “narrating”and consists of such practices as dramatizing, plot-ting, and moralizing. There is, then, considerablescope for future research to explore whether andhow practices bundle into different forms of socialordering in other risk assessment and managementcontexts and, if so, what the implications are for thenature and extent of discursive work needed tostabilize or destabilize meanings.

In addition to illustrating and theorizing howpractices and discourse intersect in the organizingprocesses through which risks are assessed andmanaged, our study also contributes to the growingbody of work adopting a performative process ap-

proach to organizing more generally. First, it showshow organizing processes bring objects into beingand construct meanings for them through the inter-play of practices and discursive work and how,depending upon the specific bundle of practicesthat is enacted, more or less contemporaneous dis-cursive work may be involved. When extensivediscursive work is carried out in the present, it ismore easily observable, as a result of which theprecarious and contestable status of meanings ismore apparent. When discursive work has takenplace in the past, the contingent status of meaningsand the body of texts holding them in place may beless evident. Our study thus illustrates how a per-formative process approach combined with a dis-cursive perspective can be used to subject all mean-ings to interrogation—to shed light on how theycome about and how they are held in place.

Second, our study provides insights into howspecific forms of social ordering bring the past tobear on the present differently through “temporal-izing” (Hodges, 2008). Practices and the ways inwhich they bundle together emanate from past or-ganizing processes in ways that are not random;they do so in particular forms of social orderingthat are familiar and relevant to actors. In addition,discursive work is carried out in the context of, andby drawing on, broader, historically situated dis-courses. In some cases, this discursive work hastaken place in the past; in other cases, it takes placein the present. Third, the study provides practicalinsights into how researchers can carry out empir-ical performative process studies. As Cooper ar-gued, researchers tend to see the world as an im-mutable system of categories and things: “Objectsof attention appear as bounded entities which existagainst a background” (2005: 1689). This tendencytoward reification makes it easier for actors andscholars alike to grasp reality, but it also hides theunderlying complexities whereby reality is made.Our study illustrates how it is possible to examinethe ways in which objects are in a perpetual state of“becoming.”

This study does, however, have limitations. First,it is a single, qualitative case study of prospectivechemical risk assessment and management pro-cesses in one country, and it compares only twochemicals within those processes. Such a focushas, however, allowed us to unpack complex pro-cesses and to subject them to a finely grained anal-ysis. Second, our main source of data has been textscontaining traces of practices rather than direct ob-servations of practices. Our focus on naturally oc-curring texts makes sense given that the specificrisk assessment and management processes westudied are highly mediated and, to a significant


extent, textual affairs (cf. Hilgartner, 1992): deci-sions are made and actions are taken on the basis oftexts that actors have accessed, read, interpreted,cited, critiqued, etc. In other words, producing,distributing, and consuming texts are important as-pects of risk assessment and management. On theother hand, our evidence rests on what actors choseto express in these texts. It is possible that obser-vations of actors as they produced the texts or ac-cess to draft versions could indicate patterns inpractices different from those we inferred from thefinal, public versions of the texts. In addition, thenorms associated with scientific text production(Knorr-Cetina, 1981) and the telling of coherentnarratives in texts designed for public consumption(Kohler Reissman, 1993) may have produced a de-gree of convergence not experienced by the actorscarrying out the practices. Ethnographic studies(e.g., Feldman, 2004) of organizing processeswould yield further and complementary insights inrelation to these issues. Finally, our study does notconsider why problematizing was more prominentin the case of BPA and much less so in the case ofVAM. Further research could explore this question,as well as the relevance and role of these bundles ofpractices in assessing and managing other types ofrisks (e.g., from nuclear power generation, off-shoreoil drilling, and emerging diseases).

These limitations notwithstanding, our studyadds to a better understanding of the processesthrough which risks are organized into existence,which is important in a risk society (Beck, 1992;Tsoukas, 1999). In theorizing how products “be-come” risk objects through organizing processes,our study contributes to orienting organization the-ory to questions of significant relevance to the pri-vate, public, and nongovernmental organizationsthat are increasingly brought together around con-temporary societal concerns about risk.

REFERENCES

Adams, M. D. 2002. The precautionary principle and therhetoric behind it. Journal of Risk Research, 5: 301–316.

Andrée, P. 2005. The Cartegena protocol on biosafety andshifts in the discourse of precaution. Global Envi-ronmental Politics, 5(4): 25–46.

Bakken, T., & Hernes, T. 2006. Organizing is both a verband noun: Weick meets Whitehead. OrganizationStudies, 27: 1599–1616.

Beck, U. 1992. Risk society: Towards a new modernity.Thousand Oaks, CA: Sage.

Beck, U. 2006. Living in the world risk society: A Hob-house Memorial Public Lecture given on Wednesday

15 February 2006 at the London School of Econom-ics. Economy and Society, 35: 329–345.

Carlsen, A. 2006. Organizational becoming a dialogicimagination of practice: The case of the indomitableGauls. Organization Science, 17: 132–149.

CEPA. 1999. Canadian Environmental Protection Act.https://www.ec.gc.ca/lcpe-cepa/default.asp?lang�En&n�26A03BFA-1. Accessed July 2010.

Chia, R. 1995. From modern to postmodern organiza-tional analysis. Organization Studies, 16: 579–604.

Chia, R. 1999. A “rhizomic” model of organizationalchange and transformation: Perspective from a meta-physics of change. British Journal of Management,10: 209–227.

Chia, R. 2002. Time, duration and simultaneity: Rethink-ing process and change in organizational analysis.Organization Studies, 23: 863–868.

Cooper, R. 2005. Relationality. Organization Studies,26: 1689–1710.

Cooper, R. 2007. Organs of process: Rethinking humanorganization. Organization Studies, 28: 1547–1573.

Cranor, C. F. 2007. Towards a non-consequentialist ap-proach to acceptable risks. In T. Lewens (Ed.), Risk:Philosophical perspectives: 36–53. London: Rout-ledge.

Dake, K. 1992. Myths of nature: culture and the socialconstruction of risk. Journal of Social Issues, 48(4):21–37.

Eccles, R. G., Newquist, S. C., & Schatz, R. 2007. Repu-tation and its risks. Harvard Business Review,85(2): 104–114.

Environment Canada. 1995/2004. Toxic substances man-agement policy. http://www.ec.gc.ca/Publications/default.asp?lang�En&xml�2EE9E1E8-1DC4-4886-93B1-D67A085FBAA3. Accessed July 2010.

Environment Canada. 2005. The Canadian Environmen-tal Protection Act (CEPA 1999) and the assessmentof existing substances. http://www.ec.gc.ca/CEPARegistry/gene_info/fact_08.cfm. Accessed July2010.

Environment Canada & Health Canada. 2008. Screeningassessment for the challenge acetic acid ethenylester (vinyl acetate monomer) chemical abstractsservice registry number 108–05-4. http://www.ec.gc.ca/substances/ese/eng/challenge/batch2/batch2_108-05-4.cfm. Accessed July 2010.

EU. 2007. European Union risk assessment report draft2007: Risk assessment vinyl acetate. CAS no. 108-05-4 EINECS no. 203-545-4, draft, June 12.

EU. 2008. European Union risk assessment report: Vi-nyl acetate. CAS no. 108–05-4 EINECS no. 203–545-4. http://ecb.jrc.ec.europa.eu/DOCUMENTS/


Existing-Chemicals/RISK_ASSESSMENT/DRAFT/R059_0807_env_hh.pdf. Accessed January 2011.

Ewald, F. 1991. Insurance and risk. In G. Burchell, C.Gordon, & P. Miller (Eds.), The Foucault effect: Es-says on governmentality: 197–210. Chicago: Uni-versity of Chicago Press.

Feldman, M. S. 2004. Resources in emerging structuresand processes of change. Organization Science, 15:295–309.

Gephart, R. P., Jr., Van Maanen, J., & Oberlechner, T.2009. Organizations and risk in late modernity. Or-ganization Studies, 30: 141–155.

Government of Canada. 2005. Assessing chemicals inCanada for risk. http://www.hc-sc.gc.ca/ewh-semt/alt_formats/hecs-sesc/pdf/pubs/contaminants/brochure/brochure-eng.pdf. Accessed July 2010.

Government of Canada. 2007a. Canada’s approach onchemical substances. http://chemicalsubstanceschimiques.gc.ca/about-apropos/canada-eng.php. Ac-cessed July 2010.

Government of Canada. 2007b. Categorization. http://chemicalsubstanceschimiques.gc.ca/fact-fait/categor_qa-qr-eng.php. Accessed July 2010.

Government of Canada. 2007c. How were substancescategorized? http://www.chemicalsubstanceschimiques.gc.ca/about-apropos/categor/how-comment-eng.php. Accessed July 2010.

Government of Canada. 2007d. Chemical substances andhumand health. http://www.chemicalsubstanceschimiques.gc.ca/fact-fait/health-sante-eng.php. Ac-cessed July 2010.

Government of Canada. 2007e. PowerPoint presentationon Canada’s Chemicals Management Plan—TheCanadian approach, June 8, 2007. http://www.ucis.pitt.edu/euce/events/policyconf/07/pdfs/glover.pdf. Accessed July 2010.

Government of Canada. 2008. Phenol, 4,4’ -(1-methyleth-ylidene)bis- (bisphenol A) CAS Registry number80–05-7. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/summary-sommaire/batch-lot-2/80-05-7-eng.php. Accessed July 2010.

Government of Canada. 2009a. Call for the nomination tothe Challenge Advisory Panel. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/panel-groupe/call-appel-eng.php. Accessed July2010.

Government of Canada. 2009b. Order amending Sched-ule I to the Hazardous Products Act (bisphenol A).http://www.gazette.gc.ca/rp-pr/p1/2009/2009-06-27/html/reg5-eng.html. Accessed July 2010.

Government of Canada. 2009c. Health Canada to host2010 World Health Organization scientific expertsmeeting on bisphenol A. http://news.gc.ca/web/article-eng.do?crtr.sj1D�&mthd�advSrch&crtr.

m n t h n d V l � 2 & n i d � 5 0 1 4 2 9 & c r t r . d p t 1 D �&crtr.tp1D�&crtr.lc1D�&crtr.yrStrtVl�2006&crtr.kw�bisphenol%2Ba&crtr.dyStrtVl�1&crtr.aud1D�&crtr.mnthStrtVl�2&crtr.yrndVl�2010&crtr.dyndVl�7. Accessed July 2010.

Government of Canada. 2009d. Overview of the Chem-icals Management Plan. http://chemicalsubstanceschimiques.gc.ca/alt_formats/pdf/fact-fait/overview-vue-eng.pdf. Accessed July 2010.

Government of Canada. 2009e. Fact sheets and fre-quently asked questions. http://www.chemicalsubstanceschimiques.gc.ca/fact- fai t/ index-eng.php. Accessed July 2010.

Government of Canada. 2009f. Government of Canadaacts to protect newborns and infants from bisphe-nol A in polycarbonate plastic baby bottles. http://news.gc.ca/web/article-eng.do?crtr.sj1D�&mthd�advSrch&crtr.mnthndVl�2&nid�462599&crtr.dpt1D�&crtr.tp1D�&crtr.lc1D�&crtr.yrStrtVl�.2006&crtr.kw�bisphenol%2Ba&crtr.dyStrtVl�26&cr t r . aud1D�&cr t r .mnthSt r tVl�2&cr t r .yrndVl�2010&crtr.dyndVl�8. Accessed July 2010.

Government of Canada. 2010a. The Government of Can-ada “Challenge” for chemical substances thatare a high priority for action. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/index-eng.php. Accessed July 2010.

Government of Canada. 2010b. Order adding a toxicsubstance to Schedule 1 to the Canadian Environ-mental Protection Act, 1999. http://www.gazette.gc.ca/rp-pr/p1/2010/2010-02-27/html/reg1-eng.html. Accessed July 2010.

Government of Canada. 2010c. Questions and answersfor action on Bisphenol A under the ChemicalsManagement Plan. ht tp : / /www.chemicalsubstanceschimiques.gc.ca/fact-fait/bisphenol-a_qa-qr-eng.php. Accessed July 2010.

Government of Canada. 2010d. What substances isCanada already managing? http://www.chemicalsubstanceschimiques.gc.ca/about-apropos/manage-gestion/already-deja-eng.php. Accessed July2010.

Government of Canada. 2010e. Deadlines associatedwith the Challenge. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/deadline-limite-eng.php. Accessed July 2010.

Government of Canada. 2010f. Chemical substances.http://www.chemicalsubstanceschimiques.gc.ca/index-eng.php. Accessed July 2010.

Grant, D., & Hardy, C. 2004. Introduction: Struggles withorganizational discourse. Organizational Studies,25: 5–14.

Hardy, C., & Maguire, S. 2010. Discourse, field-configur-ing events, and change in organizations and institu-tional fields: Narratives of DDT and the Stockholm


Convention. Academy of Management Journal, 53:1365–1392.

Hardy, C., Palmer, I., & Phillips, N. 2000. Discourse as astrategic resource. Human Relations, 53(9): 7–28.

Hardy, C., & Phillips, N. 1999. No joking matter: Discur-sive struggle in the Canadian refugee system. Orga-nization Studies, 20: 1–24.

Hardy, C., & Phillips, N. 2004. Discourse and power. In D.Grant, C. Hardy, C. Oswick, & L. Putnam (Eds.), Thehandbook of organizational discourse: 299–316.London: Sage.

Health Canada. 2008a. Health Canada responds to con-cerns raised about bisphenol A in canned food.http://www.hc-sc.gc.ca/ahc-asc/media/nr-cp/_2008/2008_84-eng.php. Accessed July 2010.

Health Canada. 2008b. Health Canada protects familieswith bisphenol A regulations. http://www.hc-sc.gc.ca/ahc-asc/media/nr-cp/_2008/2008_167-eng.php. Accessed July 2010.

Heracleous, L., & Barrett, M. 2001. Organizational changeas discourse: Communicative actions and deepstructures in the context of information technologyimplementation. Academy of Management Jour-nal, 44: 755–778.

Hilgartner, S. 1992. The social construction of risk ob-jects: Or, how to pry open networks of risk. In J. F.Short Jr. & L. Clarke (Eds.), Organizations, uncer-tainties, and risk: 39–51. Boulder, CO: Westview.

Hodges, M. 2008. Rethinking time’s arrow: Bergson, De-leuze and the anthropology of time. AnthropologicalTheory, 8: 399–429.

Hood, C., & Jones, D. 1996. Preface. In C. Hood & D. Jones(Eds.), Accident and design: Contemporary de-bates in risk management: xi–xiii. London: UCL.

Hutter, B., & Power, M. 2005. Organizational encoun-ters with risk: An introduction. In B. Hutter & M.Power (Eds.), Organizational encounters with risk:1–32. Cambridge, U.K.: Cambridge University Press.

IARC. 1995. Dry cleaning, some chlorinated solvents andother industrial chemicals. IARC Monographs on theEvaluation of Carcinogenic Risks to Humans, 63:443–465.

Jasanoff, S. 1998. The political science of risk perception.Reliability Engineering and Systems Safety, 59(1):91–99.

Keller & Heckman. 2007. Key aspects of Canada’s Chem-ical Management Program and Challenge Program.http://www.khlaw.com/Files/2556_Marraspese_KeyAspectsofCanada.pdf. Accessed July 2010.

Klinke, A., & Renn, O. 2001. Precautionary principle anddiscursive strategies: Classifying and managingrisks. Journal of Risk Research, 4(2): 159–173.

Knorr-Cetina, K. D. 1981. The manufacture of knowl-edge: An essay on the constructivist and contextualnature of science. Oxford, U.K.: Permagon.

Kohler Reissman, C. 1993. Narrative analysis. London:Sage.

Kunreuther, H., & Slovic, P. 1996. Science, values andrisk. Annals of the American Academy of Politicaland Social Science, 545: 116–125.

Lawrence, T., & Phillips, N. 2004. From Moby Dick toFree Willy: Macro-cultural discourse and institu-tional entrepreneurship in emerging institutionalfields. Organization, 11: 689–711.

Leveson, N., Dulac, N., Marais, K., & Carroll, J. 2009.Moving beyond normal accidents and high reliabil-ity organizations: A systems approach to safety incomplex systems. Organization Studies, 30: 91–114.

Levy, D. 1997. Business and international environmentaltreaties: Ozone depletion and climate change. Cali-fornia Management Review, 39(3): 54–71.

Lupton, D. 1999. Risk. London: Routledge.

Maguire, S. 2004. The coevolution of technology anddiscourse: A study of substitution processes for theinsecticide DDT. Organization Studies, 25: 113–134.

Maguire, S., & Ellis, J. 2009. The precautionary principleand risk communication. In H. D. O’Hair & R. L.Heath (Eds.), Handbook of risk and crisis commu-nication: 119–137. New York: Routledge.

Maguire, S., & Hardy, C. 2006. The emergence of newglobal institutions: A discursive perspective. Organ-ization Studies, 27: 7–29.

Maguire, S., & Hardy, C. 2009. Discourse and deinstitu-tionalization: The decline of DDT. Academy ofManagement Journal, 52: 148–178.

Malenfant, R. 2009. Risk, control and gender: Reconcil-ing production and reproduction in the risk society.Organization Studies, 30: 205–226.

Meriläinen, S., Tienari, J., Thomas, R., & Davies, A. 2004.Management consultant talk: A cross-cultural com-parison of normalizing discourse and resistance. Or-ganization, 11: 539–564.

Miller, K. 2009. Organizational risk after modernism.Organization Studies, 30: 157–180.

Mythen, G. 2008. Sociology and the art of risk. SociologyCompass, 2: 299–316.

Nayak, A. 2008. On the way to theory: A processualapproach. Organization Studies, 29: 173–190.

Phillips, N., Lawrence, T., & Hardy, C. 2004. Discourseand institutions. Academy of Management Review,29: 1–18.


Pidgeon, N., Hood, C., Jones, D., Turner, B., & Gibson, R.1992. Risk perception. In Royal Society Study Group(Eds.), Risk: Analysis, perception and manage-ment: 89–134. London: Royal Society.

Power, M. 2005. The invention of operational risk. Reviewof International Political Economy, 12: 577–599.

Power, M. 2007. Organized uncertainty: Designing aworld of risk management. Oxford, U.K.: Univer-sity of Oxford.

Power, M., Scheytt, T., Soin, K., & Sahlin, K. 2009. Repu-tational risk as a logic of organizing in late moder-nity. Organization Studies, 30: 301–324.

Rayner, S. 1992. Cultural theory and risk analysis. In S.Krimsky & D. Golding (Eds.), Social theories of risk:83–115. London: Praeger.

Reach for Unbleached! 2008. Comments on Draft riskassessments and draft risk management scope forbatch 2 Challenge substances. http://www.cen-rce.org/CMP/pdf/Batch%202%20RFU%20comments.pdf. Accessed July 2010.

Reed, M. 1998. Organizational analysis as discourse anal-ysis: A critique. In T. Keenoy, D. Grant, & C. Oswick(Eds.), Discourse and organization: 193–213. Lon-don: Sage.

Renn, O. 1992. Concepts of risk: A classification. In S.Krimsky & D. Golding (Eds.), Social theories of risk:53–79. London: Praeger.

Renn, O. 2008. Risk governance: Coping with uncer-tainty in a complex world. London: Earthscan.

Robichaud, D., Giroux, H., & Taylor, J. R. 2004. Themetaconversation: The recursive property of lan-guage as a key to organizing. Academy of Manage-ment Review, 29: 617–634.

Schatzki, T. R. 2001. Introduction: Practice theory. In K.Knorr-Cetina & E. von Sevigny (Eds.), The practiceturn in contemporary theory. London: Routledge.

Schatzki, T. 2006. On organizations as they happen. Or-ganization Studies, 27: 1863–1873.

SCHER. 2008. Risk assessment report on vinyl acetate:Human health part. CAS no. 108-05-4 EINECS no.203-545-4. European Commission Directorate-Gen-eral for Health & Consumers. November 17.

Scheytt, T., Soin, K., Sahlin-Anderson, K., & Power, M.2006. Introduction: Organizations, risk and regula-tion. Journal of Management Studies, 43: 1331–1337.

Slovic, P. 1998. The risk game. Reliability Engineeringand System Safety, 59: 73–77.

Smallman, C. 1996. Risk and organizational behaviour: Aresearch model. Disaster Prevention and Manage-ment, 5(2): 12–26.

SRA. 2003. Risk analysis glossary. http://www.sra.org/resources_glossary. Accessed August 2005.

Sullivan-Taylor, B., & Wilson, D. C. 2009. Managing thethreat of terrorism in British travel and leisure or-ganizations. Organization Studies, 30: 251–276.

Tansey, J., & O’Riordan, T. 1999. Cultural theory andrisk: A review. Health, Risk & Society, 1: 71–90.

Thomas, R., Sargent, L., & Hardy, C. 2011. Managingorganizational change: Negotiating meaning andpower-resistance relations. Organization Science,22: 22–41.

Topal, C. 2009. The construction of general public inter-est: Risk, legitimacy, and power in a public hearing.Organization Studies, 30: 277–300.

Tsoukas, H. 1999. David and Goliath in the risk society:Making sense of the conflict between Shell andGreenpeace in the North Sea. Organization, 6: 499–528.

Tsoukas, H. 2005. Afterword: why language matters inthe analysis of organizational change. Journal ofChange Management, 18: 96–104.

Tsoukas, H., & Chia, R. 2002. On organizational becom-ing: Rethinking organizational change. Organiza-tion Science, 13: 567–582.

UNEP. 2006. Dubai declaration on international chem-icals management. http://www.chem.unep.ch/unepsaicm/mainstreaming/default.htm. Accessedon 12/8/2010.

Van de Ven, A. H., & Poole, M. S. 1990. Methods forstudying innovation development in the Minnesotainnovation research program. Organization Sci-ence, 1: 313–334.

Van de Ven, A. H., & Poole, M. S. 2005. Alternativeapproaches for studying organizational change. Or-ganization Studies, 26: 1377–1404.

Vinyl Acetate Council. 2009. Health and environment(excerpt from the “Vinyl acetate safe handlingguide”). http://www.vinylacetate.org/health.shtml.Accessed July 2010.

Weick, K., & Sutcliffe, K. 2001. Managing the unex-pected: Assuring high performance in an age ofcomplexity. San Francisco: Jossey-Bass.

Winch, G., & Maytorena, E., 2009. Making good sense:Assessing the quality of risky decision-making. Or-ganization Studies, 30: 181–203.

Wynne, B. 1992. Risk and social learning: Reifica-tion to engagement. In S. Krimsky & D. Golding(Eds.), Social theories of risk: 275–297. London:Praeger.

Zinn, J. O. 2008. Social theories of risk and uncertainty.Oxford, U.K., & Malden, MA: Blackwell.


APPENDIX

In-Text Citing Notation for Official Challenge Documents

In-text CitingNotation Source

BPA-CAP, 2008 Government of Canada, 2008. Chemicals Management Plan–Challenge Advisory Panel. SummaryReport: March 28, 2008. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/panel-groupe/summary-sommaire/200803-eng.php. Accessed July 2010.

BPA-DSAR, 2008 Government of Canada, 2008. Draft Screening Assessment for Phenol, 4,4’—(1-methylethylidene)bis-(80-05-7).

BPA-FSAR, 2008 Government of Canada, 2008. Screening Assessment for the Challenge: Phenol, 4,4’ —(1-methylethy/idene)bis-(Bisphenol A), Chemical Abstracts Service Registry Number 80-05-7. http://www.ec.gc.ca/ese-ees/3C756383-BEB3-45D5-B8D3-E8C800F35243/batch2_80-05-7_en.pdf. Accessed July 2010.

BPA-RMS, 2008 Government of Canada, 2008. Risk Management Scope for Phenol, 4,4’—(methylethylidene)bis-(Bisphenol A), Chemical Abstract Service (CAS) Registry Number: 80-05-7. http://www.ec.gc.ca/substances/ese/eng/challenge/batch2/batch2_80-05-7_rms.cfm. Accessed July 2010.

BPA-SPC, 2008 Government of Canada, 2008. Summary of Public Comments Received on the Government of Canada’sDraft Screening Assessment Report and Risk Management Scope on Bisphenol A (CAS RN 80-05-7),http://www.ec.gc.ca/ese-ees/6486D85F-0DC9-4E53-AFCD-E5242B153A44/batch2_80-05-7_pc_en.pdf.Accessed July 2010.

VAM-CAP, 2008 Government of Canada, 2008. Chemicals Management Plan–Challenge Advisory Panel. SummaryReport: From the meetings held October 20 and 27, 2008. http://www.chemicalsubstanceschimiques.gc.ca/challenge-defi/panel-groupe/summary-sommaire/200810-eng.php. Accessed July 2010.

VAM-DSAR, 2008 Government of Canada, 2008. Draft Screening Assessment for Vinyl Acetate (108-05-4), Government ofCanada.

VAM-FSAR, 2008 Government of Canada, 2008. Screening Assessment for the Challenge: Acetic acid ethenyl ester (VinylAcetate Monomer) Chemical Abstracts Service Registry Number 108-05-4. http://www.ec.gc.ca/ese-ees/E41E17F4-59C5-44CC-94F3-155CE6094238/batch2_108-05-4_en.pdf. Accessed July 2010.

VAM-RMS, 2008 Government of Canada, 2008. Risk Management Scope for Acetic acid ethenyl ester (Vinyl AcetateMonomer), Chemical Abstract Service (CAS) Registry Number: 108-05-4. http://www.ec.gc.ca/substances/ese/eng/challenge/batch2/batch2_108-05-4_rms_en.pdf. Accessed July 2010.

VAM-SPC, 2008 Government of Canada, 2008. Summary of Public Comments Received on the Government of Canada’sDraft Screening Assessment Report on Vinyl Acetate (CAS No. 108-05-4). http://www.ec.gc.ca/ese-ees/59EC93F6-2C5D-42B4-BB09-EB198C44788D/batch2_108-05-4_pc_en.pdf. Accessed July 2010.

Steven Maguire ([email protected]) is the found-ing director and chair of the Marcel Desautels Institute forIntegrated Management, as well as associate professor ofstrategy and organization, in the Desautels Faculty of Man-agement at McGill University. He received his Ph.D. fromHEC-Montreal. He focuses on institutional and technologi-cal change, seeking to understand the processes throughwhich technologies enter and exit the economy and howthese processes are influenced by the strategic behaviors ofnonmarket actors (e.g., NGOs, scientists, and governmentorganizations) in addition to market ones (e.g., firms andtheir customers). His current research focuses on risk,chemicals management, and green chemistry.

Cynthia Hardy ([email protected]) is a Mel-bourne Laureate Professor of Management at the Uni-versity of Melbourne, codirector of the InternationalCentre for Research on Organizational Discourse, Strat-egy & Change, honorary professor at Cardiff BusinessSchool, and a Fellow of the Academy of the SocialSciences in Australia. She received her Ph.D. from theUniversity of Warwick. Her research interests includea discursive perspective on institutional entrepreneur-ship and institutional change, power and politics inorganizations, organizational discourse theory, andcritical discourse analysis.


ORGANIZING PROCESSES AND THE CONSTRUCTION OF …...Smallman, 1996). Despite some positive...

Documents

Transcript of ORGANIZING PROCESSES AND THE CONSTRUCTION OF …...Smallman, 1996). Despite some positive...