collective sensemaking about reading: how teachers mediate … · 2018-10-16 · granted status as...
TRANSCRIPT
Educational Evalutation and Policy AnalysisSummner 2001, Vol. 23, No. 2, pp. 145-170
Collective Sensemaking about Reading: How Teachers MediateReading Policy in Their Professional Communities
Cynthia E. CoburnUniversity of Pittsburgh
Recent research on the relationship benveen instructional policy and classroom practice suggests thatteachers interpret, adapt, anlCd even transform policies as they put them itito place. This paper extendsthis line of research. using at in-depth case study of one California elementaty school to examine theprocesses by which teachers coonstruct and reconstruct multiple policy messages about reeading in-struction ini the context of their prqtessional commtnunities. Drawing primarily on institutional andsensemaking theory, this paper puts forth a mocdel of collective sensermaking that focuses on the waysteachers co-construct understandings of policy messages, make decisions aboiut which messages topursue in their classrooms, and negotiate the technical anzd practical details of implementation in con-versations with their colleagues. It also argues that the nature and structutre offormal networks andinformnal alliances amnong teachers shape the process, with implications ftr ways in which messagesfrom the poliey environment influence classroom practice. Finally, the paper explores the role schoolleaders play in shaping the sensemaking process.
Studies of school refonn have often asked how agiven reform impacted schools, or how teachershave implemented a particular policy. Yet someresearchers have suggested that rather than policyinfluencing teachers' practice, it is more likelythat teachers shape policy. That is, teachers inter-pret, adapt, and even transform reforms as theyput them into place (Cohen & Ball, 1990; Tyack& Cuban, 1995; Weatherley & Lipsky, 1977). Asimportant as this insight is, there has been littlesystematic research into the processes by whichsuch interpretations and adaptations occur. How-ever, a promising strand of research points toteachers' professional communities as importantsites for this meaning making, highlighting theways in which local teacher communities can formpowerful microcultures (Little & McLaaughlin,1993; McLaughlin, 1993; Siskin, 1994) thatmediate environmental pressures (McLaughlin& Talbert, in press; Spillane, 1999; Talbert &McLaughlin, 1994).
This article examines the processes by whichteachers construct and reconstruct multiple mes-sages about reading instruction in the context of
their professional communities. Previous researchexploring the ways teachers make sense of policymessages has tended to focus solely on individualinterpretation (Cohen & Ball, 1990; Jennings,1996; Spillane & Jennings, 1997; Weatherley &Lipsky, 1977). A few studies, however, havebegun to move beyond individual interpretation,providing evidence that individuals make senseof policy messages in conversation with their col-leagues (Spillane, 1999; Hill, 1999) and in waysthat are deeply situated in broader social, profes-sional, and organizational contexts (Lin, 2(000;Spillane 1998; Yanow, 1996). This paper buildson and extends this line of research. For a year,I followed teachers in one urban Califomnia ele-mentary school as they sought to improve theirreading instruction, focusing on the ways they col-lectively negotiated pressures and interpreted andadapted messages from the environrnent. To un-derstand how this process unfolded, I have drawnprimarily on the theoretical and empirical work ofinstitutional and sensemaking theory. I argue thatthe nature and structure of formial networks and in-formal alliances among teachers play a powerful
145
Coburn
role in shaping the sensernaking process and ulti-mately the kind of sense that is made. The processthat emerged was one in which teachers, in con-versations with their colleagues in formal and in-formal settings9 co-constructecd understanidings ofmessages from the environment, made decisionsabTout which messages to pursue in their class-roon, and negotiated technical and practical de-tails of implementation.
Reading instruction in California provides afertile context for this study because the state is inthe midst of its seco nd major shift in reading pol-icy in only 15 years. Following the rise of themovemlent toward "whole language" instructionin the late 1980s, there has been tremendous ac-tivity and controversy since the mid- i 990s as thestate and the profession have moved towardsomewhat different versions of what is frequentlycalled a "balanced" approach to reading instruc-tion, This historical moment has thrown into re-lief the often subtle and taken-for-granted processof interpretation and adaptation. Uncovering theprocess by which teachers reconstruct policymessages in their professional communities iscrucial. It contributes to our understaniding of therelationship between instructional policy andclassroom practice. It mnay also provide insightinto ways policy can create conditions for sense-making in schools that enable teachers to engagemiessages fromi the environment in ways that en-couErage them to challenge their assumptions andcoontinue to inprove their practice over timie.
Bridging Institutional anidSensernaking Theory
Most studies concerned with the relationshipbetween policy and instructional practice focus onhow teachers and schools respond to a single pol-icy or a network of related policy initiatives (see,for example, Odden, 1991). Yet in spite of recentmiovement in state and local policymiaking towardcongruence (Fuhrman, 1993; Smith & O'Day,1991), many schools cotntinue to find themnselvesresponding simultaneously to multiple initiatives.Furtherm-ore, many schools and individual teach-ers are involved in reforTm efforts outside theformal policy systerm-various school reformmnodels, work with professional developmelntproviders., coursework at universities, and teachernetworks, to name just a few. Thus, teachers oftenfind themiiselves confronted with multiple mes-sages about reading-normative pressures about
146
how they "should'9 teach, belief systems aboutteaching and learning, and specific teaching prac-tices--fromn a wide range of sources.
Institutional theory provides powerful toolsfor understanding the comiiplicated relationshipbetween schools and their multifaceted environ-ment. At root, institutional theory is a cultural ap-proach. It emphasizes how norms and culturalconceptions about appropriate reading instruc-tion are constructed and reconstructed over timne,carried by individual and collective actors, andemnbedded within policy and governance struc-tures (Scott, 1995; Scott, Menidel & Pollack.1996). Institutional theorists suggest that, mes-sages in the environment shape patterns of actionand belief within schools through regul(ative
means, as they are incorporated into formal pol-icy: through normantive means, as teachers feelpressured to adopt certain approaches to inain-tain legitimacy; and through cognitive means, asreading beliefs and practices attain taken-for-granted status as the natural or commn-on-senseway to do things (Scott, 1995). Messages aboutreading are thus "carried" by policy at all levelsof the system and through reform programs,teacher professional organizations, assessmentsystems, textbooks and other materials, profes-sional development, community expectations,and individual and coliective actors. As a carrier,formal policy (at the state, district, and school lev-els) is only one of many mechaniisms by whichmessages about reading coine into schools.
Traditionally, institutional theory has been ap-plied ratlher narrowly to schools. Most studieshave focused on the influence of the institutionalenvironment on1 school structures and organiza-tion. These studies have largely failed to investi-gate empirically the connections between theenvironment and teachers' work in classrooms,relying instead on earlier theorizing that suggestedthat schools decouple struictural changes from theinternal workings of the classroomi (Meyer &Rowan, 1977, 1978). Furthermore, many institu-tional studies of public schools have presented asimpli fied "outside-in" model in which bel ief sys-temns, nortmis, and practices originate in the envi-ronment and flow into schools (see. for example,Cuban, 1988: Malen & Ogawa, 1988: Rowan,1982). This approach neglects the dynanmic rela-tionship between the environment and schools' in-ternal social processes, failing to account for theevidence-highlighted by educational historians
Collective Sensernaking about Reading
such as Tyack and Cuban (1,995)-that teachersactively mediate nonns, belief systems. and prac-tices that have diffused from the institutional en-vironrment, socially constructing and reconstnict-ing them as they put them into place in their owncontexts.
I draw on sensemaking theory for guidance inexploring how teachers have adapted, adopted,combined, or ignored messages and pressuresabout reading instruction in their professionalcommunities, and how these deliberations haveshaped classroom practice. Sensemaking theoristssuggest that school and classroom culture, struc-ture, and routines result, in part, from "micro-momentary actions" by teachers and other actorsin the school (Porac, Thomas, & Baden-Fuller,1989). Action is based on how people notice orselect information from the environment, makemeaning of that information, and then act on thoseinterpretations, developing culture, social struc-tures, and routines over time (Porac et al., 1989;Weick, 1995). The meaning of information orevents-in this case, mnessages about reading-isnot given, but is inherently problematic; individu-als and groups must actively construct under-standings and interpretations. They do so by plac-ing new information into preexisting cognitiveframewvorks, also called "worldviews" by sometheorists (Porac et al., 1989; Vaughan, 1996;Weick, 1995).1 Thus, teachers notice new rues-sages and construct understandings of themthrough the lens of their preexisting practices andworldviews (EEPA, 1990; Jennings, 1996;Spillane & Jennings, 1997; Spillane, 1999).
Sensemaking is not solely an individual affair,but is social in two important respects. First, it iscollective in the sense that it is rooted in social in-teraction and negotiation. People make sense ofmessages in the environment in conversation andinteraction with their colleagues, constructing whatI call "shared understandings"-organization- andworkgroup-specific cultuLre, beliefs, and routines-along the way (Porac et al, 1989; Vaughan, 1996).Second, sensemaking is social in the sense that itis deeply situated in teachers' emnbedded contexts.Norms and routines of organizational subunitssuch as departments or workgroups (Siskin, 1994;Spillane, 1998; Vaughan, 1996), organizationalvalues and traditions (Lin, 2000; Porac et al.,1989), and broader professional culture (Barley,1986: Spillane, 1998; Vaughan, 1996) provideanother lens through which teachers make sense
of new messages, shaping the range of appropri-ate responses and structuring priorities. Further-mnore, these embedded contexts shape sense-making processes by influencing patterns ofsocial interactions (influencing who is talkingwith whom about what) and shaping conditionsfor sensemnaking (Vaughan, 1996).
This article focuses on the collective aspects ofsensemnaking with attention to the way that it issituated in and shaped by teachers' broader em-bedded contexts. As such, it makes three key con-tributions to earlier work in both sensemakingtheory and cognitive perspectives on policy im-plemeentation. First, by studying teacher sense-making in action over a long period of time. I amable to unpack some of the main components ofthe process itself, Earlier research has providedconvincing evidence for the claims that teachersreconstruct policy ideas through their preexistingbeliefs and practices. This study provides an elab-orated account for how that process unfolds, de-veloping a model that identifies key subprocesses.Second, the paper highlights the central role ofteachers' for)mal networks and informal alliancesin the patterns and outcomies of collective sense-making. Finally, the study brings school-levelleadership into the teacher sensemaking equation,exploring the connections between actions by theprincipal and teacher leaders and the nature andcontent of teachers' sensemaking.
Methods
To capture senseraking, I used a qualitativecase study approach. a primary strategy for doc-umenting organizational processes as they un-fold (Yin, 1984). Focusing on a single case al-lowed for the depth of observation necessary tocapture the subtle and iterative process by whichteachers constructed and reconstr-ucted messagesfrom the environment through social interaction.Although not generalizable, the in-depth obser-vation made possible by the single case providesthe opportunity to generate new hypothesesor build theory about sets of relationships thatwould otherwise have remained invisible (Hart-ley, 1994). I relied primarily on sustained obser-vation (Barley, 1990) and in-depth interviewing(Spradley, 1979), supplementing these strategieswith document analysis. The case study schoolwas selected because it is an urban school inCalifornia involved in an ongoing effort to im-prove reading instruction. Stadele Elementary2 is
147
Cobuen
exceptionally racially diverse, and the vast mna-jority of the student body lives in poverty; manyare also English [anguage learners.3 I focused onteachers in the early grades-especially first andsecond grade-because this is the level of read-ing instruction that has been at the center ofpolicymiakinig and debate in recent years.
At the time of the study, Stadele Elementarywas in its secontd year participating in a school re-form program that used whole-school inrquiry as alever for instructional change. As such, it receivedfuniding to design its own whole-school "focusedeffort" and to assess its progress using an inquiryprocess. The reformn effort was led at the schoolsite by a leadership teamn composed of the princi-pal, a half-time reformn coordinator (a releasedclassroom teacher), four classroorn teachers rep-resenting primnary and upper elemrientary grades,and three resource teachers (two Reading Recov-ery teachers4 and the coordinator of parent out-reach). The school designed their focused effort todevelop standards and grade-level indicators forreading, and to assess stutdent progress towardmeeting the standiard in reading comprehension.They used the mnajority of their funding to releasegrade-level teams for orne-half day a month towrite standards, develop assessments, and exam-ine their reading instruction. G3rade-level effortswere coordinated during bimonthly, fLull facultymeetings after sch0ool and during pupil-free pro-fessional development days funded by the stateand district. The school also used reform fundingto send teams of teachers (ofte-n members of theleadership teamn) to external professional devel-opment related to reading instruction and as-sessment. All of these meetings and professionaldevelopment opportunities provided amnple oc-casions to observe sensemaking in action.
Observatlon of inforrnal and formnal teachermeetings form-ned the centerpiece of data collec-tion activities. Over the course of the 1998-1999school year, I spent more than 130 hours observ-ing teacher conversations during formal meet-ings and professional developmcnt. I spent themiajority of timre observing grade-level mneetingswith the first- and second-grade groups (14.5 and16 hours respectively) and full faculty meetings(43.5 hoturs). But I also observed in-schnool pro-fessional development, select m-eetings of othergrade-level grouips, anid required district profes-sional development. in addition, I spent signifi-cant time with the leadership teamn, observing
their meetings (13.5 hours) and attending externalprofessional development with them (I1 .5 hours).In addition to formal miieetings and professionaldevelopment. I observed counitless hours of infor-mal conversations during lunch, before and afterschool, and in the hallways. See Table Al in theAppendix for complete information on mneetingobservations. During observations of formal andinfoirmal conversations, I paid attention to the na-ture and coontent of imnessages about reading thatteachers cante into contact with, the content oftheir conversations with one another, the nature ofiniteraction, as well as e-vidence of teachers' world-views and practice. While miost observations (for-m-al and informal) were typed up as field notes, ona few occasions I taped and transcribed key ineet-ings that I was unable to attend.
In order to capture teachers' worldviews, de-scriptions of their practices, and perspectives onthe reform process, I supplemented observationswith semistructured interviews with classroomteachers, resource personnel, and the principal. Iadopted a strategy that combined breadth (initialinterviews with nearly all first- and second-gradeteachers, members of the leadership team, and re-source teachers) with depth (intensive interviewsand observations with a subset of teachers andresource staff). In all. I conducted 57 interviewswith 18 classroom and resource teachers, inter-viewing some teachers as many as 12 times. I alsointerviewed the principal three times. Interviewslasted from 45 minutes to three hours. Nearly allinterviews were audiotaped and transcribed. SeeTable A2 in the Appendix for additional infor-mation about the distribution of interviews.
To uncderstand the relationship between con-versations in teacher meetings and teachers' read-ing practices, I observed reading instruction inselect teachers' classrooms. Using informationgarnered from first-round interviews, I selectedfor observation teachers who represented the fullrange of approaches to reading instruction usedby teachers in the early grades of the school. Iconducted 106 hours of observation in the class-rooms of five first- and second-grade teachers andthree Reading Recovery teachers. I structured ob-servations to spend fuIll days in a teacher's class-room for several days in a row in the first half andthen again in the second half of the year. Spend-ing a full day, rather than simply observing thetimie period in the morning a teacher designatedas "reading instruction," was importanit for two
Collective Senisemaking about Reading
reasons. First, mIost teachers actually engaged ina range of reading-related activities throughout theday. Second, shadowing a teacher for an entire dayrather than dropping in for a bounded time periodproved important for observing teachers' ad hocand informal conversations with their colleagues.Observing classrooms for several days in a rowprovided a sense of the flow and continuity ofinstruction in the near term, and doing observa-tions at two different times of the year allowedfor insight into change over time.5 See Table A2for more information about the distribution ofobservations. Finally, I relied on record dataand interviews with district personnel, state per-sonnel, and local professional development pro-viders to understand the nature of the readingenvironment.
Data collection and analysis occurred simulta-neously throughout the study year (Miles & Hu-berman, 1994; Strauss & Corbin, 1990). 1 usedinitial codes to identify emerging themes andhighlight areas for additional data collection.After all data were collected, I used NUDISTqualitative data analysis software to code obser-vations of forrnal and informal meetings in threeways. First, 1 coded all meeting data using codesdescribing the nature of teachers' interaction withmcessages from the environment. Here, I devel-oped codes inductively through iterative coding(Miles & Huberman, 1994; Strauss &. Corbin,1990). I began with codes that described, with lit-tle interpretation, the nature of teachers' inter-action with messages from the environment. Bygrouping together categories and using the con-stant comparative method (Strauss & Corbin,1990), 1 moved to progressively higher levels ofabstraction until I ended up with the followingcodes: constructing understanding, gatekeeping,negotiating technical/practical details, in-facing,and out-facing. Second, I coded the content of theconversations, creating codes such as "'new text-book series," "district standards," and "readingcomprehension strategies" that correspondedwith the specific messages that teachers came intocontact with throughout the year. Once the datawere coded in this manner, I was able to createlongitudinal records of teacher interaction in agiven group around a particular message or set ofnmessages, thus tracing the course of teachers'conversation about them over time. Third, I codedall meeting data for evidence of teachers' world-views, descriptions of reading practices, and evi-
dence of groups of teachers' shared understand-ings. I coded interview data using similar codes,supplemented with codes for teacher's history,school context, and the environment (includingdistrict, state, and the larger debate about readinginstruction). Finally, I coded classroom observa-tions using codes for curriculum, instruction, as-sessment, and environment.
In developing an account of the processes in-volved in collective sensemaking, the group orcollectivity was a key analytic unit. Thus, I en-gaged in data analysis that looked across allformal and informal groups, but also comparedresponses of groups as they interacted with spe-cific miessages (keeping in mind that there wasoverlapping group membership and that not allgroups interacted around all messages). To de-velop a model of the key subprocesses involvedin collective sensemaking, I engaged in system-atic, inductive coding (described earlier) of alloccasions on which formal and informal teachergroups interacted with a range of messages fromthe environment. Further analysis involved re-coding a category at a finer level of detail. For ex-ample, I recoded the subprocess "gatekeeping"by doing a content analysis of the reasons teach-ers made gatekeeping decisions. To understandthe relationship between teachers' worldviews,practices, and shared understanding and thesesubprocesses, I created and then compared lon-gitudinal records of conversations on particularmessages across different and, at times, overlap-ping groups (Strauss & Corbin, 1990; Miles &Huberman, 1994).
To understand the relationship between collec-tive sensemaking and changes in classroom prac-tice, I compared the content and nature of con-versation in formnal and infornal meetings withevidence of teachers' responses to messages fromclassroorn observations, at times supplementingthis information with evidence of practice from in-terviews and meeting obser-vations. As it becameclear that the relative heterogeneity of worldviewspresent in a group, whether interaction occurred informal or informal settings, and the structure ofactivities in formal mneetings were important inboth the nature of teacher interaction and implica-tions for classroom practice, I created additionaldata displays to confirm these pattems, alwaysbeing alert to disconfirning evidence (Miles &Hubermnan, 1994). Finally, I explored the role ofrefonm leadership by tracing the language, actions,
149
Cobhurn
and decisions leaders mnade with the content andprocesses of interaction in teacher glroups.
Several methodological features of the studyensure that the patte-rns reported here representpatterns present in the research site. These strate-gies include intensive immersion at the researchsite (Eisen-hart & Howe, 1992; Lofland & Loflantd,1995), systematic sampling of occasions for sense-making (Miles & Huberrnan, 1994). efforts toexplore countervailing evidence (Miles & Hu-bermaan, 1994), systematic coding of data (Strauss& Corbin, 1990: Miles & Huberinan, 1994), andsharinig findings with key informanits at the schoolsite and inicorporating their insights into the finalanalysis (Eisenhart & Howe, 1992; Miles &Hubernnan, 1994).
Messages about Reading in the Environment
In 1995, the state of California launched areading initiative that moved away from-l earlierstate policy, clharacterized by somne as "wholelanguage," toward a position that has comine to beknown as the "balanced approach" to reading in-struction. Backed by trem-aendous material re-sources, the reading initiative has been excep-tionally comiiprehensive and wide-reaching. Sincethe publication of the Task Force report EveyChild a Reader (Californil Department of Edu-cation, 1995), which ooutlined in broad terms thevision of the balanced approach to reading in-struction, the state has passed five bills appropri-ating funds to purchase instructional materials,as well as four bills to provide professional de-velopment for teachers and district leadership onapproaches to reading instruction that emphasize"explicit and systematic" approaches to phonics.It has adopted new reading series, developedstate standards and curTiculuni framneworks, andadopted a new statewide assessment system. Fi-nally, the state has provided futnding to collegesof education to move preservice teacher educa-tion in this direction and instituted a state examfocused on reading instruction that new teachersmrust now pass in order to becontie credentialed(California State Board of Educationw, 1999). Thesestate efforts have had enormous public visibilitythanks to extensive media coverage that has tendedto lambaste the earlier policy position and call fornew approaches to teaching reaLling emphasizingphonics.
Although the state policy, with its extensiveresources and comprehensive reach, created a
150
strong priority for improving reading instruictionand pushed the debate in a particular direction,the specitic messages about reading instructionembedded wvithin that policy were not always thesame ones to reach Stadele Elementary. First.state policy is often reinterpreted and reshapedby policy makers at every level as it works itsway through a system to schools (Hill, 1999;McLaughlin, 1991b; Pressm-nan & Wildavsky,1984: Spiliane, 1996). In this case, the districeand, to a lesser extent outside professional de-velopmetit providers, played key roles in recon-structing state policy guidelines as they devel-oped district policy and provided professionaldevelop,-ment to schools.6 Second, there can be aconsiderable time lag between policy mnaking atthe state level and when messages reach schools(Tyack & Cuban, 1995). Although state policyhas consistently promoted a "balanced approach"to reading instruction since 1995, the meaning ofthat term has shifted from the relatively ambigu-ous and relatIvely uncontested notion of "bal-ance" in Eveiy Child a Reader to the much mnorespecific and arguably more controversial rendi-tion represented in the 1998 Reading/LanguageArts Framiework-for CIalifrnia Public Schools(California Department of Educationi, 1999). Thistr ansition has narrowed the scope of whtat is con-sidered appropriate reading instruction in statepolicy. However, mnany of the messages enteringStadele Elemlentary School in the 1 998-99 schoolyear were representative of the broader definitionof balance in earlier policy making. includingmnany approaches that have since boee explicitlycontradlicted by recent state policy.7 And crucially,as institutional theoty has emnphasized, schoolsand teachers come into contact with mressagesfrom a wiide range of sources, many of wlhich areoutside the fornal policy system-other refornefforts, professionial development, preservice ed-ucation, and connectionis with colleagues insideand outside of school. Soietinies, these messagesare aligned with state policy, sometimes they arenot, and sometimes they simply go beyond thebounds of areas addressed by policy.
For all these reasons, I take as the point of de-parture for analysis lnot the state-level policy butall of the heterogeneous messages that actuallycame into the school from all sources during the1998-99 school year. During that year, StadeleElementary School came into contact with nies-sages about reading fromn three key sources: tune
Collective Sensemaking about Reading
district as it reconstructed state policy mandates,participation in the school reformn program, andindividual teachers' coinections to the environ-inent. The district disseminated its own standardsdocument accompanied by a core curriculum withmodel lessons and assessments, provided profes-sional development on the new reading seriesto all of its elementary school teachers, and in-creased emphasis on new state standardized testsby requiring schools to use them to guide their pri-orities for school improvement and measure theirprogress. The school's reform effort also served todraw district policy-especially the standards,district assessment kits, and the reading series-more deeply into the school, as the leadershipteam structured reform activities around thesedocuments. Beyonid district policy, the school'sleadership team brought in messages about read-ing from the broader network associated with thereforrn effort, including resolrces from externalprofessional development and feedback from part-nerships with other schools. Finally, the team,drawing on the school's own resources, had theschool's Reading Recovery teachers provide twoprofessional development sessions on giving andanalyzing a reading assessmeent called runningrecords.
In addition to responding to messages carriedby district policy or the school's reform effort,teachers often actively reached out for messagesabout reading. These messages and the ways theyentered the school were quite diverse, from moreexperienced teachers drawing on resources fromolder, more traditional approaches, to teachersdrawing on experiences in preservice that sup-porte.d previous state policy (whole language), toteachers bringing in mnaterials they learned aboutat a conference with a reading researcher whosework helped shape current state policy.
The environment for readiing is multifacetedand multilayered. Teachers came into contactwith many, many messages about reading fromdiverse sources stretchinig far beyond formal pol-icy streams. State policy played an important roleby providing funding and normative pressure thatcreated key opportunities for increased activityand policy making at lower levels of the system(Spillane, 1996). It also pushed other sources ofmessages in the environment in the direction ofa "balanced approach." However, though mostof the messages coming into Stadele Elemen-tary were supportive of a "balanced approach"
broadly conceived, many were not well alignedwith the narrower conception of "balance" foundin current state policy. Furthermore, because ofthe complexity of the environment and the multi-ple routes into the school, still other messagesrepresented older approaches to reading that, al-though out of favor in the professional and policyenvironment, were still carried into conversationsby individual teachers.
Collective Sensemaking about Reading
When confronted with new messages. theearly-grade teachers at Stadele Elementary tendedto turn to their colleagues to make sense of them.In contrast to traditional images of isolated teach-ers who have few conversations about teachingand learning (Goodlad, 1984; Lortie, 1975), teach-ers at Stadele Elementary talked with one anotherabout their practice in multiple settings. The re-form effort built on a long history of faculty andgrade-level meetings, providing funding for addi-tional meeting time and shifting the balance ofconversations away from administrative matterstoward issues of assessment and instruction. Be-yond formal settings, however, most teacherssought out like-minded colleagues to talk abouttheir classrooms. These more informal conversa-tions happened in an ad hoc manner before school,after school, and during lunchtime as teachersasked each other questions, discussed their stu-dents, and shared resources. Conversations inthese settings represented collective sensemakingto the degree that teachers made sense of newmessages about reading in ways that involved so-cial interaction, negotiation, signaling, and com-munication with colleagues, regardless of whetheror not a group of teachers came up with a singledecision or interpretation. What is more importantis that the teacher sensemaking happened in andwas influenced by this social interaction. In thissection, I first put forth a model of the process bywhich teachers adopted, adapted, and at timestransformed messages about reading during inter-action with colleagues in formal and informal set-tings. I then argue that collective sensemaking-and, ultimately, the influence of messages fromthe environment on classroom practice-is shapedby two factors: (I) the patterns of interactionamong teachers, specifically who is talking withwhom in what setting, and (2) the character ofconversation, specifically the extent to which
151
Coburn
conversations are structured to provide conditionsfor engagement and reflection.
The Sensemaking Process
in both formal and informal groups, teachers'interaction with messages from the environmentfollowed a pattern similar to that observed in thesensemaking literature. Specifically, however, Iidentify three clear subprocesses that characterizeand facilitate collective sensemaking: (1) con-structing understanding through interpersonalinteraction, (2) gatekeeping, and (3) niegotiatingtechnical amd practical details. Eachi of these threesubprocesses was itself influenced by teachers'woridviews, preexisting practices, and sharedunderstandings. That is, teachers brought theirworldviews and preexisting ways of teachingreading to interactions with their colleagutes. And,as teachers worked together over time, they de-veloped shared understandings.t Bv influencing
the subprocesses, teachers' worldviews and prac-tiees, as well as groups of teachers' shared under-standings, shaped, what teachers selected, em-phasized, interpreted, and ultimnately brought intotheir classrooms. Figure I provides a representa-tion of this process.
Construicting Understanding throughInterpersonal Interaction
When teachers came into contact with newrnessages about reading instruction, they oftenspent time with their colleagues constructing anunderstanding of what the messages meant. Somemessages, of course, seemed self-evident, wereregistered with little conversation or exploration,and seemed to be integrated searnlesslv into cur-rent understandings. But other messages requiredquite a bit of conversation for teachers to makemeaning of what otherwise would simply havebeen words on a page or a description of an in-
Messages from the Environment
/
FIGURE 1. Conceptual modlel of the senisemaking proress.
152
reading practices
woridviews
sharedunderstandings
constructingunderstandings
gatekeeping
technical/practicaldetails
Collective Senseinaking about Reading
structional approach. For example, during onegrade-level meeting, teachers were told to assessthemselves on a rubric of assessment practicesprovided to the school by the reforn program.The rubric had a particular point of view aboutappropriate assessment practices for reading in-struction, including a heavy emphasis on usingassessment to inforn practice on a continuingbasis, and including students and families as part-ners in assessmenit. To use the rubric as a planningtool, grade levels needed to come to an under-standing about what these concepts meant. Onegrade-level group had a long conversation aboutwhat it meant to have farnilies as partners in theassessment process:
T eacher I: I think it's when we lei parentsknow what we are doing.
Teacher H: That's progress, not assessment.I think it means that we have parents helping uscomne up with the rubric.
Teacher G: If we had parents as partners, wewould have parents in here with us today doingthis work with us.
leac er H: A partnership would mean hav-ing parents helping us determinie how we aregoing to do assessment.
Teacher J: Parents are supportive of whatwe're doing.
Teacher H: That's two-way communication;it's not settinig up an assessment system. Par-ents aren't in here deciding what spe'ling wordswe will have on the spelling tests.
Here, teachers put forth and modified differentinterpretations of what it might mean to havefamilies as partners. In a sense, each of these in-terpretations franmed the concept in differentways by linking the language of the rubric "fam-ilies as partners in the assessment process" to dif-ferent familiar frameworks-parent involvementas teachers sharing infornation versus partner-ship as work-ing together side by side. As the con-versation unfolded beyond the excerpt presentedabove, teachers framed, refranied, and elibhoratedtheir various conceptions until they were able tocome up with a rendition that provided concep-tual "hooks" that allowed teachers in the group tolink this idea with what they knew and believedabout interaction with parents. Snow and his col-leagues (1986) describe this process as "framealignment." In this case, the group constructed ashared understanding that having families aspartners meant having parents working side by
side with them on all assessment tasks (as op-posed to sharing the results of assessmeent withparents or involving them in some less intensiveway throughout the process). This conception offamilies as partners, once constructed, then per-sisted over time. It served as a jumping-off pointfor teachers' decisions about where to placethemselves on the rubric and how to plan for im-provement. And teachers in this group referredback to this language and framning when the issueof their relationships with parents came up inlater conversations, within and otutside of fonnalmeetings.
The individual and collective worldviews rep-resented in a group played a key role in shapingthe process by which teachers constructed sharedunderstandings. This phenomenon was illus-trated tnost clearly when teacher groups withvery different worldviews and practices con-structed different understandings of the samer.essages. In one example, using a process simi-lar to the one described above, teachers in onegroup came to a shared understanding of what itmeant to use assessment to inform instruction ona continuous basis, another item on the rubric.They framed the concept using their understand-ing of reading instructitn as structured accordingto a particular sequence of skills. They con-cluded: "[Using assessment to iniform instruc-tion] is the skill work .... We plan learning cen-ter work based on the skills they need to knowand it goes in a particular sequence." For thisgroup, then, using assessment to inform instruc-tion meant knowing where in the sequence achild was and planning lessons accordingly.
In contrast, another group of teachers believedthat skills should he taught in response to theneeds of children rather than in a set sequence.Based on that belief, these teachers were able toconstruct an understanding of the same conceptin a different way-as developing lessons in re-sponse to the particular needs of the student, nomatter what the sequence. During the meeting,one teacher summarized the group's understand-ing in the following way: "So we're saying it'songoing observation of students ... yon are con-stantly looking at what they are doing and whenyou see something they are having trouble with,you plan a minilesson." Here, the configurationof groups was critical. Because different groupswere composed of teachers with very differentworldviews and shared understandings, teachers
153
C'oburn
in the group framed the idea of using assessmnentto inform instruction in very different ways. Sig-nificantly. there was little overlap in the range ofinterpretations offered in the course of coniversa-tions in these two different groups.
(Gatekeeping
During the course of the year, teachers wereconfronted with an enormous number of mes-sages about reading instruction. Obviously, not alltnessages could be ineorporated, and so teachers'professional commrtunities played a crucial gate-keeping role. Once teachers constructed an under-standing of what a given message was about, theyeither engaged with the idea or approach, or theycismissed it. In this way, teachers essentially se-lected sonie messages in and selected others out.In conversations with their colleagues, teachersrejected messages fromr the environment for arange of reasons, many of which were linked withtheir worldviews or shared understandings. Unlessotherwise noted, each of the followinlg reasonswas observed across all teacher groups.
Does not appily to their grade level. Notions ofgrade-level appropriateness exert powerful nor-mative pressure. In the course of conversation,groups of teachers rejected messages about whatthey "should" be doing from the district standards,professional developm.ent, and assessments basedon their sense that the particular approaches werenot appropriate to their grade level. However.teachers in different groups often constructedsomewhat different conceptiois of what was ap-propriate for a given grade level. For examtiple,after attending professional development on read-ing comprehension strategies, two teachers in thefirst grade decided during a lunchtirne coniversa-tion that higher order cornprehension strategieswere inappropriate for first graders. They thenmade gatekeeping decisions on this basis. In con-trast, two other teachers-a first- and a second-grade teacher-came to see thie strategies as ap-propriate to both first grade anad second grade anddecided to meet after school to talk about ways tobring the approach into both of their classrooms.
Yaoo diffic/ult for tieir students. During fon-nal andinformal conversations, teachers drew on theirsomnetimes extensive experience to assess whetherstudents in their classroom would be able to hian-dle a particular activity or master a particular skill
154
or strategy. In interaction, teacher groups drewfrom their individual experienices and expecta-tions to construct and abridge their shared con-ception of'what is realistic for "kids in this school"to know and he able to do. In so doing, they oftenoverruiled professional developers or new districtstandards as suggesting soniething they consid-ered too difficult for their students.9
Philosophically opposed. Individual teachersoften brought activities and resources to formaland informal conversations with colleagues. Inggroups of teachers with divergent views-espe-cially grade-level groups-resources brought byindividuals were sometimes rejected by others inthe group because of ph1ilosophical oppositioII.For example, in one grade-level group, oneteacher's repeated suggestions that they use areading comprehension assessment that usedstory excerpts and multiple-choice questionswere rebuffed by the rest of the group, which sawthe assessment as "too traditional." Interestingly,what was considered philosophically problem-atic varied fromn group to group. Had this teacherbeen a participant in the other grade-level group,it is more likely, given the comnposition of thegroup, that her stiggestions and resources wouldhave been incorporated into group activities.
Comitpletely outside the bounds of comtprehen-sibilirv. Teachers also rejected messages aboutreading in conversation with their colleagues be-cause they involved approaches that seemed in-conceivable-so far outside of the bounds of whatthey saw as appropriate-that they were not evenworth considering. In one example, early-gradesteachers attended a district professional develop-nient day on the new reading series run by trainedteacher leaders in which one teacher leader sug-gested individualized instruction as a possible wayto teach reading. Later, as teachers in one grade-level group reflected on the professional develop-ment, they camne to agree that individualized in-struction was entirely inappropriate. One teacherexpressed this shared sentiment as she said to hercolleagues, "How can you possibly teach readingwithout putting kids in groups? That's crazy@"Again, different groups of teachers found differ-ent messages about reading inconceivable, de-pending upon the worldviews represented in thegroup and the nature of the shared understandingsthey had developed over time.
Collective Sensernaking about Reading
Other reasons were influenced by what teacherswere already doing in their classrooms, or whatSarason calls "preexisting regularities" (Sarason,1971):
Doesn't 'fit." Most teachers in the school hadwell-developed and quite complicated structuresfor teaching reading in their classrooms-waysof organizing time and children, materials thatthey used, and kinds of activity structures. Dur-ing formal and infornal conversations, teachersdiscussed whether or not and how new materialsor approaches would fit with their individual andshared approaches. When these conversationsdid not produce a way to work new messages intopreexisting programs, teachers selected thoseapproaches out.
Unmainageable. Teachers rejected messages orapproaches when they jointly decided that it wasnot possible to successfully implement theem inthe classroom becautse of time and other con-straints, burdens of record keeping, or concernfor behavior management issues.
Finally, teachers rejected m essages when thereweren't adequate knowledge resources availableto the group.
Dicta 't feel they understood it. When there wasno one in a group who understood particular lan-guage in the standards, particular approaches of-fered in professional developpment, or particularassessment instrLuments, the group tended to skipover them. For examnple, while putting togethertheir assessment program, teachers in both grade-level groups came across assessments that re-flected newer understandings of phonics andphonemic awareness that they didn't understand.Teachers in the group weren't sure why theywould want to assess such a thing and they couldnot figure out how to adimlinister the assessment.It perhaps comes as no surprise that without un-derstanding the assessments, and without findingthe resources for greater understanding withintheir proximal communities, these groups soughtout different assessments to include in their as-sessment system.
Negotiating Technical and Pra-ctical Details
Sensemaking theorists talk about the ways inwhich interpretation is extemalized in action(Porac et al., 1989; Weick, 1995). For teachers atStadele Elementary, this link between meaning
making and action was far from straightforward.For those mnessages that they did not dismiss dur-ing gatekeeping, teachers talked with colleaguesto work out the technical and practical detailsinvolved with moving an idea, a particular ap-proach, or teaching materials into the classroom.This attention to the technical and practical notonly represented a significant percentage of teach-ers' conversations with one another, but was alsocrucial in order for teachers to translate abstractideas emrlbedded in messages from the environ-nent into concrete action in the classroom. This
translation can be particularly challenging givenboth the complexity of reading instruction and thecomplexity of the classroom. Teachers' conversa-tions about technical and practical details rangedfronm coming to understand how to use particularaspects of the textbook series in the context of aparticular teacher's reading program; to workingout timing, format, and record keeping for assess-inents; to exploring different ways to group stu-dents; to discussing what kind of paper to use fora particular activity. The conversations tended tobe iterative, brought up again and again over time,sometimes after one or more teachers had tried apractice or used some matei'ais in the classroom.
As with other parts of the sensemaking process,how teachers translated messages about readinginto classroom practice was shaped by world-views, preexisting practice, and shared under-standings. Preexisting practice and the structuralconstraints at the school played an especially im-portant role in framiing how teachers wo3rked outthe technical and practical details. For example,both grade-level groups spent a lot of time figur-ing out how to create time to give individualizedassessments given classroom managemenit con-cerns (what will the other children be doing?),scheduling of other school events that cut intoreading time, and the pressing need for test prepactivities.
In addition, conversations about technical andpractical details provided yet another occasion forinterpretation and meaning making. In one grade-level group, teachers' different mental modelsfor how children leam to read created points ofdisagreement as they figured out the technicaldetails of administering running records. Run-ning records are a form of assessment that pro-vides insight into both a child's reading process(what strategies the child is using to decode) anda child's reading level (what books the child can
i 55
Cohumn'
read on an "instructional" level) (Fountas & Pin-nell, 1996). In deciding upoen a common procedurefor administering the running records, the teach-ers got into an argumnent about whether to use astory with pictures or to use text excerpts withoutpictures:
Teacher G: I know. We could do runningrecords on this (referring to a series of leveledtext excerpts developed for another purpose in-cltided in the district assessment compendium).
Teacher 1: 1 want to use something with pic-tures because that is an important part ot reading.
Teacher F: I agree. That's what I don't likeabout [the standardized test]. There are no pic-tures and that is an important part of reading.
Teacher G: But this [the text excerpts with-out pictures] is words and souunding out words.It is really interesting to see which kids can lookat it and really soundi out words.
T'eacher G's preference for using text withoutpictures was rooted in an instructional mnodelstressing the primary importance of attention tophonics (graphophonic cues) in decoding print-a view that was evidenit in interviews and obser-vations of her classroomn. In contrast, Teachers Iand F saw reading as involving a broader range ofstrategies for decodinig, including attention to pie-ture cues-also reflected in their intet'views andclassroom observations.1 1 ' Ultimately, TeacherG's view prevailed and the group ended up doingrunning records on text excerpts without pictures.In this instance, the process of itnteraction and ne-gotiation unfolded in a manner somrewhat differ-ent frotn earlier examnples. Teachers in this groupfound themselves unable to construct a way ofadministering the running records that accom-modated teachers' very different underlying be-liefs about what it is important to assess in read-ing. Instead, they constructed a conmpromiseposition in which they agreed to admiiinister therunning records three times a year for school as-sessment purposes without pictures, while ad-ministerinig them however they wanted the restof the year. Teacher I later explained that after ayear of conversations made challenging by thediverse woridviews represented in the group, shedecided it was easier to give in for grade-level as-sessments and just continue to do her own thingin the classroom. For this teacher, group inter-action shaped her routines, but not her underlyingview of the reading process. In contrast, the de-cision did influence the assessment practices of
156
two teachers in the group with less well devel-oped views on the nature of reading instruction.After giving running records without pictures forthe grade-level assessment, these teachers re-ported that they coontinued to use running recordsin their class in this mnanner onl an ongoing basis.But perhaps the main point here is that whileworking out the technical details of implemnenta-tion, conceptions among these teachers of howchildren leam to read shaped how they put run-ning records into practice.1 '
While I have separated the three subprocessesinivolved in collective sensemaking for the sakeof conceptual clatity, in practice few conversa-tions marched through these steps in a straight-forward manner. Rather. sensemaking in formaland infornal settings was both highly iterative andrecursive. Teachiers returned to issues over andover throtughout the year, modifying their inter-pretations, reconsidering technical and practicalco,ncerns, and often making new gatekeepingdecisions as they catne into contact with addi-tional messages fromii the environment or exper-irmented with new approaches or materials in theirclassrooms.
Factors Affecting Sensemaking
While the basic process of sensemaking re-mained similar among and between differentgroups of teachers, patterns of teacher interactionand the conditions for the conversation were cru-cial for shaping if and how messages from theenvironment influenced teachers' worldviewsand classroom practice.
Patterns (of linteraction: Makitng Different Sentseoj the Same Messages
Whomn teachers talked with in what settingmattered because teachers in different groupsoften impade different sense of the same messages.Teachers worked together both in formal settingswhen they were asked to work on tasks related tothe reform effort, and in informal settings whenthey chose to talk with one another about theirclassrooms. It was significant that teachers foundthemselves working with very different people indifferent settings. In formal settings, teacherswere most often grouped by grade level.'2 Butwhen teachers chose whom to talk with on theirown, they tended to self-select into informalnetworks with similar worldviews and ap-proaches."3 For example, teachers in the first
Collective Sensemnaking about Readi7zg
grade stratified into two separate groups-agroup of older, more experienced teachers whofavored direct instruction and workshect activi-ties, and a group of new teachers who favored fa-cilitative kinds of teaching and interdisciplinary,active leaming activities. In another example, newteachers tended to reach out to one of three Read-ing Recovery teachers for guidance, because thisteacher had a reputationi for innovative and pro-gressive teaching when she had her own class-room at the school several years earlier. In aschool that tended to have pedagogical splitsalong generational lines, new teachers gravitatedto an older, experienced teacher whom they per-ceived to have compatible worldviews and ap-proaches, as well as expertise in early reading in-struction.' 4 There were also a few teachers in theearly grades who had few, if any, ties with otherteachers in informal settings. In the absence ofstrong collegial ties, sensemaking appeared tohappen in a more isolated fashion.
Over time, as teachers worked closely withone another infomially, their practice and world-views became increasingly similar within eachgroup through a process of reciprocal influenice.The tendency to seek out like-minded teachers ininformal settings created a situationi of pedagog-ical diversity in formal settings and relative ho-mnogeneity in informal settings. It also meant thatthe differences between informal groups wereoften greater than those between formal ones.
Because different groups were composed ofdifferent teachers with contrasting worldviews,preexisting practices, and shared understandings,teachers in different formal and informnal groupsinterpreted and actualized messages from the en-vironment in substantively different ways. Forexample, drawing heavily on their preexistingreading practices and assumptions about the ap-propriate way to teach reading, teachers in differ-ent informal networks made different sense of thereading series and ended up using it in entirelydifferent ways. After considerable discussion andsome experimentation, one pair of teachers, in aclear example of gatekeeping, rejected the read-ing series entirely because they felt it did not fitwith the way that they structured reading instruc-tion in their reading groups. Rather than intro-ducing sequenced phonics and comprehensionskill work prior to the story, as these teacherswere used to doing with earlier reading series, thenew series emphasized activating students' back-
ground knowledge prior to the story and thenteaching skills on an as-needed basis in the con-text of the story. Having decided in conversationwith one another that "you can't teach kids to readusing [the new reading series]," the two teacherscontinued to use the old reading series for theirreading groups.
In contrast, two teachers in another informalgroup who taught reading in the context of largerthematic units, saw the textbook as a source ofstories to use with their themes rather than a cur-riculunm to follow. As such, the particulatr lessonstructure advocated by the textbook was less im-portant than the nature of the stories. They de-cided to use the textbook, but shared wisdomwith each other about how to pick and choosestones that fit with their themes, bringing the newreading series into each of their classrooms in away that adapted it to their preexisting programrather than guiding their program by the readingseries. Not all teachers made sense of the newreading series in conversation with their col-leagues. But those who did showed evidence ofthe mutual influence in their approaches to andunderstanding of the textbooks, A trio of teach-ers who worked together ended up using the newreading series, following it in order. Two otherteachers used the supplementary set of phonicsreaders, but not the main textbook.
The degree of heterogeneity in a group alsoplayed a role in the ways teachers interacted withmessages from the environment. Informal set-tings, because of their pedagogical homogeneity,were more supportive, but also more conserva-tive. As teachers with similar worldviews andpractices interacted to make sense of a message-especially those teachers who had a long historyof working together-they relied Tnore on sharedunderstandings to construct understanding. Thus,they did not engage in the kind of framing and re-frainiing that tended to surface, question, and attimes shift assumptions. But while the diversityof approaches and perspectives in formal settingstheoretically offered the opportunity for teachersto challenge one another's worldviews and learnfrom each others' approaches to reading instruc-tion, when the range of worldviews and practicesin a group was too large, teachers had difficultycommunicating across them. This happened inboth grade-level groups around different issues.When it happened, coming to a shared under-standing of messages from the environment, mak-
157
Cioburn
ing gatekeeping decisions, and working out thetechnical details of implemientation was chal-lenging, involving much negotiating, arguing, andconvincing in conversations that stretched overmiany meetings. As was the case in the runningrecord example, when groups could not constructan understanding or negotiate technical and prac-tical details in ways that made sense to all in-volved, and when there was external pressure fora joint solution, groups constructed negotiatedresponses that reflected the construction of mnoredominant members of the group. As illustratedearlier, individiuals who did not subscribe to groupdecisionis often did not mnake changes in theirclassrooms.
Conditions Supporting Deep Engagenent:Jn-ft'cing anet Out-fJlcing Con versations
Conversations differed markedly in the degreeto which they fostered deep engagement. Thisengagement was signiificanit because without it,teachers were more likely to summarily dismissmessages without exploring themn and were lesslikely to engage with their colleagues and mes-sages from the environment in ways that causedthem to question their worldviews, practices, andshared understandings. Activities in formal set-tings at Stadele Elementary did not always pro-vide the coniditions for this deep engagement.Activities associated with the reform effort wereoften highly structured, w/ithi "assigniments" fromi
the leadership team including guiding questionsanld products related to the reform effort. Whilesome assignmrents seemed to engage teachers(especially those directly related to developinigand scoring assessmiients), teachers often foundit difficult to see the coninection between otheractivities and the work they were doing in theirclassrooms (especially school-level planning ac-tivities, assigned discussion topics, and self-assessments regarding process). In the words ofone teacher: "What we're supposed to do isn'tnecessarily the direction we choose, so some-timies it's hiarcd to get a fire under us." Or teachersdid not understand why they were being directedto do certain things. For example, during onemeeting, grade-level grotips were instructed to"reflect' on questions raised about the school'sreading instruction by outside "critical friends"who visited the school to observe for a day. Thequestionis, while somewhat different fromr activ-ities they had performned in the past. seemed verysimilar to the group:
158
Teacher G: [reading the question I "How isreading comprehension assessed at your gradelevel?" Haven't we answered this already?
Teacher .K: I am so bored with this
Collective Senseniaking about Reading
brought into their classrooms. An analysis link-ing in-facing conversations in informal settingsto classroom observations, for example, providesevidence of a pattern of multiple, consistent, in-cremental changes in classroom practice relatedto teachers' interaction with their colleaguesaround messages about reading (see Coburn,2001). In addition, a minority of in-facing con-versations caused teachers to question their world-views, practices, ancd shared understandings. Forexample, when one early-grades teacher becamedissatisfied with the way she organized her classduring reading groups, she began talking with asecond teacher about alternative strategies, aconversation that stretched over many months.The second teacher described an organizationalsystemn for rotating students through readinggroups and learning centers that she had adaptedfrom professional development. This approachgrouped students homogeneously for readinggroups and heterogeneously for lear-ning centers.The first teacher was hesitant because the approachwas complex and she felt that homogeneousgrouping was iniequitable. However, during thecourse of repeated conversations, the secondteacher was able to answer the first teacher's con-cerns by arguing that using homogeneous group-ing for reading groups was equitable becausestudents learn best when they are reading attheir instructional level. The second teacher alsoworked with the first teacher to figure out how tomake the approach fit with the particular contextof her reading program and students. Thus, inconversation with the second teacher, the firstteacher rethought and subsequently reshaped thefundamental organization of her classroom dur-ing reading instruction. During the course of thestudy year. I identified three such occasions onwhich teachers questioned their preexisting prac-tices and worldviews in fundamental ways dur-ing conversations in informal settings.
Somiie activity structures or assignments in for-mal settings did elicit in-facing conversations. Ofthe 27 observed activities, 15 provided condi-tions for in-facing conversations, But as in infor-mnal settings, conversations in which teachers in-teracted with messages from the environment inways that caused them to question their world-views or practices were rare: I identified five in-stances in which teachers began to rethink theirpractices in more than incremental ways duringformal meetings. Most of these instances hap-
pened as teachers jointly scored student assess-mients. In one example, one grade-level group'sdeliberations while scoring student assessmentsled them to question taken-for-granted assump-tions about the nature of reading comprehension.Over the course of several meetings they asked:"What is reading comprehension anyway?" "Howdo story maps promote reading comprehension?""How should we address fnultiple problems andsolutions that children identify in the story?""What kinds of teacher questions really help chil-dren develop itnproved reading comprehension?"When questions were raised that teachers did nothave answers for in their collective knowledgeand experience, they tended to turn to availableexternal sources (such as the district standards andthe reading series) and engaged these ideas inways that shaped their shared understandings,worldviews, and reading practices. Teachiers ques-tioned their worldviews and practices in formalsettings when activities were linked to classroompractice, when teachers were not so far apart inworldview and practices that they could still com-nunicate with one another, and when they did notfeel pressured to get through tasks quickly to meetimpending deadlines.
Throughout the sensemaking process, theni,teachers in Stadele Elementary worked with theircolleagues in various ways to adapt, adopt, com-bine, or reject messages about reading from theenvironmnent. What sense teachers ultimatelymade--the way in which they constructed un-derstanding, made decisions to select some ap-proaches and not others, and worked out techni-cal and practical details necessary to enact theinterpretation-was deeply shaped by whomteachers were working with and the conditionsfor conversation. The conditions for conversa-tion were crucial because without the opportu-nity for deep engagement, without the time andstructure to delve into and construct an under-standing of messages fromn the environtnent andfigure out ways to integrate new practices orideas into the complex world of the classroomi, itwas unlikely that messages about reading touchedclassroom practice. Furthernore, because theway in which messages influenced the classroomwas heavily framed by teachers' worldviews, pre-existing practices, and shared understandings,whom teachers were making sense with shapedhow messages influenced teachers' beliefs andpractices.
159
Cohurn
Perhaps because teachers most often foundconditions for engagemnent in informal settingsamyong colleagues with similar worldviews andpractices, they often responded in ways that re-affirmed and reenacted preexisting practices andbeliefs. But at the saime time, new shared under-standings that resulted from negotiation, the dia-logue between colleagues that at times surfaceddeeply held assumptions, and the new approaches,ideas and materials that were incorporated in theclassroomii also reshaped worldviews., practices,and shared understandings. In this way, somemessages from the environnment did influenceclassroom practice. albeit in ways that were moreincremental than that imaginIed by most policy-nakers.
Role of Reform Leadership
Senseniaking in teacher nicrocommunities wasdeeply situated in the larger school cotntext. In par-ticular, the principal and others in the school whotook leadership in the reform effort played a keyrole in shaping the sensemaking process by inlflu-encing where sensemaking happened. by bringingin and privileging certain mitessages about readingand not others, by being strong voices in the con-struction of untlerstanding, and by structUFring thecollaboration in formal settings.
Shaping Where Sensenmakinig Happened
As described earlier, teachers in Stadele Ele-mentary School spent a lot of time talking withtheir colleagues about teaching and learniing inboth formal and informal settings. However,in spite of calls for increasing opportunities forteacher collaboration in schools (Little. 1982;Louis & Marks, 1998; McLaughlin, 1993; New-mann & Associates, 1996). not all schools haveas many formnal opportunities for teachers to worktogethier or the culture of collegiality that fosteredthe high level of informnal interaction found at thisschool (Hargreaves, 1994: Little, 1990). To theextent that refon-n leadership at the school shapedthe opportunities teachers had to talk with eachother about issues of teachitng and learning, it canbe argued that they shiaped the degree to whichthis sensemaking happened in professional com-munities rather than on an inidividual level.
The currenit levels of collegiality and collabo-ration at Stadele Elementary were fostered overmany years as currenit and former principals
160
placed a priority on teachers' work together andencouraged a culture of collegiality outside offormal settings. T'eachers with long histories inthe school remembered a time when teacherswere more isolated from one another, with few ifany forrmal opportunities to collaborate. This pat-tern changed in the imid- I 980s with the arrival ofthe previous principal arnd the advent of partici-pation in a court-ordered desegregation programthat mandated a high degree of professionaldevelopment. With the principal's leadership,teachers began to meet for the first timiie in formalgrade-level groups to work on joint projects andbegan to share experiences in professional de-velopment that encouraged them to talk aboutteaching and learning. The school began to builda culture of collegiality that eventually movedbeyond formal settings to widespread iniformalsettings. Even when that principa l left the schooland some of the grade-level projects subsided,this culture of collaboration in informal settingsflourished and continued. When the school be-camne itnvolved in the new refonn project, the lead-ership team again placed a priority on teacher col-laboration,jb usinlg a substantial percentage of theirreform money to fund released time for teachersto work with one another on refonn goals.
Brirnging in anid Privileginig Messagesfrtom the Environment
Refortn leadership played a major role in bring-ing in and privileging some messages from the en-vironmnent and filtering others out. The principal,and to a lesser extent other members of the lead-ership tearm, had greater access to messages fromthe environment than most classroom teachers.They were the ones who attended district meetingsand networking events associated with the reformeffort. Similarly, the Reading Recovery teacherson the leadership team were connected with boththe broader Reading Recovery network and read-ing leadership at the district. Through these con-nections, the principal and other reform leaderslealrned about new materials, belief systems, andideas and rnade key decisions about which mes-sages to pass on to the staff and which to filter out.In one example, members of the leadership teamattended professional development on reading in-struction that featured speakers on topics includ-ing explicit, systematic phonics instruction andreading comprehension strategies instruction (thatalso promoted teaching phonics in context). The
Collective Sensemaking about Reading
leadership team decided to provide professionaldevelopment to the staff only on what they learnedabout reading comprehension strategies, filteringout the messages about decoding. Furthermnore,when they provided the professional develop-ment, they highlighted some approaches and ideas(e.g., the six strategies for reading comprehensionused by proficient readers highlighted in the pro-fessional development) but not others (e.g., theparticular pedagogical approach to teaching thesesix strategies).
But beyond bringing in messages from the en-vironment, the principal and leadership team alsofocused schoolwide professional developmentand designed meeting activities in ways that priv-ileged certain messages over others. For example,they made the decision to have professionaldevelopnment on running records after seeing somany other schools using running records to as-sess their students. Similarly, the principal andleadership team structured activities in grade-leveland whole group mneetings in ways that focusedattention on district standards and assessmentresources. For example, while putting togethergrade-level assessment programs, teachers wereprompted to select assessmnents that were part ofthe core curriculum and the reading series. 17 As aresult of the way the leadership team privilegedcertain messages. teacher groups focused theirattention in their work together on such things asrunning records, district assessments, and the stan-dards rather than other myriad messages that caneinto the school.
Framing Messages
Some voices are more influential than othersin the process of constructing the meaning ofmessages about reading from the environment(Isabella, 1990). The principal and the reform co-ordinator framed messages from the environ-ment in ways that shaped conversation not onlyin whole group meetings when they were pres-ent, but also in grade-level meetings when theyweren't. For example, the principal repeatedlyfrarned the meaninig of the new reading series asone of many tools that could be used to meet thestandards, positioning the standards as the cur-riculum rather than the textbook. In one meetingshe said: "There is the core curriculum and thestandards, and then there is the adoption [readingseries]. With the adoption, fthe district] picks thematerials that are closest to the standard, but no
publisher can be the be-all or end-all.... Thereare other options.... You're supposed to teachto the stancdards. The adoption is one way to sup-port it, but not the only way." Teachers repeatedthis construction of the meaning and significanceof the reading series vis-a-vis the standards asthey discussed with each other how to use thereading series. In a sense, the principal's con-struction authorized teachers to use the readingseries in a wide range of ways, or not at all, as longas they geared their instruction to the standards.And, as discussed earlier, there was enormousdiversity in the way teachers came to use the read-ing series, including several teachers in each gradelevel who chose not to use the reading series at all.In addition to the reading series, the principal andreform leadership played a key role in framingother messages from the environment: the natureof standards, the purpose of doing assessment, theappropriate response to standardized-test pres-sure, and approaches to teaching reading compre-hension strategies.
Structuring Collaboration
Finally, reform leadership played a key role instructuring activities in ways that created very dif-ferent opportuniities for teachers to engage deeplywith messages from the environment. Reformleaders often structured collaboration by creatingactivities or guiding questions for teachers to fol-low in grade-level groups or the whole school. Asdiscussed earlier, some activity structures seemneddeeply linked to classroomi practice, while othersseemned more distant. Some activities encouragedand enabled teachers to challenge their own andothers' worldviews and approaches to reading in-struction, while others did not. Whether reforTnleaders were able to make fornal settings produc-tive places for teachers to engage with ideas fromthe environment in ways that made a difference inthe classroom depended on three factors. First, itdepended upon the degree to which reform lead-ers were able to create au hentic activities withenough flexibility that teachers were able to drawconnections between the activities and their class-roomns. Seconid, it depended upon their ability toprovide enough support to encourage teachers toengage with infonnation that at times challengedpreexisting ways of doing things. Finally, formalsettings were productive places when there wasenough time for teachers to revisit and rethink newpractices.
161
Coburiz
Reformii leaders played an important role in
shaping the sensemaking process as it unfolded inboth formal and informal teacher groups. They
created conditions that enabled teachers to talkwith one amotlher and, at times, created the struc-
tures that encouraged teachers to critically exam-ine their worldviews and practices. But beyondthat, reform leaders also played a powerful role in
shaping the direction of sensemaking by creating
pathways that brought particular messages from
the environment into teacher conversations, by
privileging certain messages over others, and by
framing interpretations in powerful ways.
Conclusion
For the last two decades, research on the rela-
tionship between policy and practice has notedthe ways that policy developed at higher levels of
the system gets reconstructed at the school level
(see McLaugblin, 1991 a, for a review). Alth:ougl
several studies have attem-npted to understand thisphenoimenon by looking at individual teachers'
practice (EEPA, 1990; Jennings, 1996: Spillane& Jennings, 1997; Weatherley & Lipsky, 1977),relatively few studies have explored the role of
teachers' professional communities (Spiianle,1999, and Hill, 1999, are notable exceptions). Yetin schools whiere teachers interact with one an-other in either formal or informal settings, these
conversations can play a powerful role in miediat-ing between messages from the environment and
what teachers bring into their classrooms.By examining the nature and content of teach-
ers' conversations with one another in one urban
elementary school, this article analyzes the pro-cess by which teachers construct and reconstructmessages from the environment in their profes-sional communities. I argue that patterns of inter-action and the conditions of conversation in for-mnal and informal settings influence the process by
which teaclhers adopt, adapt, comnbine, and ig-nore messages from the enivironment, mediatingthe way messages from the environment shapeclassroom practice. Other researchers have ar-gued that professional communities are not uni-tary (McLaughlin & Talbert, in press; Siskin,1994; Talbert & McLaughlin, 1994). This studygoes further, highlighting the important role ofteacher interaction outside of formnal organiza-tional structures-interaction that can be highlyinfluential in the ways that teachers make sense ofmessages from the environment. Recognizingthat the teacher cotimmunity is multifaceted and
162
dynamic is important because teachers in differ-ent formal and informal communiities can makedif'ferent sense of the same thing. Individualteachers' worldviews and preexisting praetices,and groups' shared understandings shapc howteachers construct understandings, select somilemessages in and others out, and negotiate thetechnical and practical details necessary to trans-late abstract miessages into concrete action. Fi-nally, this up-close examination also shows thatformal policy is only one of many sources of mries-sages and pressuires about reading that teacherscome into contact with. To focus on fornial pol-icy alone is to misrepresent all that teachers areresponding to and grappling with as they work toimprove thcir practice.
This portrayal of collective sensenmak-ing-both as playing a key role in shaping the waysmessages about reading actually become a part ofclassroomn practice aund as a complex process un-folding differently in many parts of the school-raises key questions for policy. From a policy-maker's perspective, it may seem that schools andteachers-in reconstructing and reinterpretingpolicy messages-are subverting the intent ofpolicy or thwarting implementationi. After all,teachers did not always make sense of messagesfroi-n the environmiient with colleagues in waysthat policymakers might have hoped (see Hill,1999, for another example of this phenomenon).But another way to look at it is that this sense-making is both necessarv and unavoidable. Pol-icy and other sources of messages about readingare by nature abstract (also see Hill, 1999, on thlispoint). Messages about reading come into schoolsin the form of ideas, inaterials, amid descriptions ofpractices. Yet teachers' work by nature involvesaction. Teachers are faced with multiple mies-sages and pressures to improve reading instruc-tion and must find ways to make meaning ofthem, translating abstract ideas into action in thecontext of their classrooms.
Conversations with colleagues may facilitatethis process. For teachers at Stadele Elementary,that the sensemaking happened with colleagues,as opposed to individually, enabled them morereadily to integrate new ideas into the highly sit-uated context of their classrooms. Work- withother teachers helped themn grapple with multipleand sometimes conflicting messages. Tt broughtaccess to greater resources and expertise. It helpedthem make decisions about which of the plethoraof messages to pursue. And, perhaps most im-
Collective Sensernaking about Reading
portant, it helped them to construct what Spillane(1 999) calls the "practical knowledge" necessaryto turn abstract ideas into somnething workable intheir classrooms.
Furthermnore, sensemaking in communities hasthe potential to provide conditions for teachers toengage with messages from the environment inways that encourage them to question their as-sumptions, challenge their frames, and continueto improve their practice over time. Throughinteraction, teachers gained access to a range ofinterpretations and ways of negotiating the tech-nical and practical details that went beyond theirown experiences and worldviews. Via the collec-tive negotiation of a range of interpretations, theyoften developed new constructions of messagesand new strategies for integrating approaches intotheir classrooms that extended, elaborated, and, ina few cases, transformed preexisting individualand group worldviews and practices.
But not all conditions for collective sense-making provided teachers opportunities to learnand grow. When conversations were structuredaround activities with little connection to theclassroom, with tine frames that did not allow forin-depth conversations, or occuiTed in groups ofteachers with such divergent worldviews andpractices that communication became difficult,teachers were often unable to engage with mes-sages in more than superficial ways. Furthennore,conversations at times encouraged a tendency forteachers to reconstruct messages in ways that re-inforced preexisting worldviews and practices, en-couraging stability of practice rather than change.
Given that the process of reconstruction maiybe a central part of moving ideas into practice, thequestion for policy is: How can policy encourageconditions for collective sensemak-ing in schoolsthat promote learning and growth? The experi-ences of teachers in Stadele Elementary offersome insight into this question.
First, policy can find ways to encourage a col-laborative culture in schools. Informal networksamong teachers are largely unacknowledged bythe policy world. Yet they have enormous poten-tial to play an influential role in teacher sense-making. It is their flexibility, their spontaneity,their voluntariness, and their situatedness thatmake informal groups such powerful and sup-portive contexts for teacher sensemaking (Har-greaves, 1994). A key question for policy, then, ishow to encourage and enable this kind of culture,without imposing the kinds of "contrived colle-
giality" (Hargreaves, 1994) that may be associ-ated with mandated collaboration. The experi-ences of Stadele Elementary suggest that the for-mal policy system canl support in-formal networksby providing funding for shared experiences ofhigh-quality sustained professional developmentfocused on teaching and learning. At StadeleElementary, these shared experiences over a longperiod of time provided teachers with a commonlanguage, common experiences, and much fod-der for conversations that extended beyond for-mal meetings into informal networks. Further-more, the experiences of Stadele Elementaryecho previous research that suggests that the prin-cipal plays a key role in setting a tone of opennessand conmrmunication and a focus on teaching andlearning that encourage a culture that movesaway from isolation toward mutual supportaround matters of instruction (Hargreaves, 1994;Louis, Marks, & Kruse, 1996).
Second, policy can find ways to foster condi-tions for in-facing collaboration in formal settings.The challenge for policy is how to create the con-ditions that enable teachers to work with one an-other in formal settings in waxs that promote thekind of in-facing conversations often found in in-formal settings, but also draw upon the power ofthe diversity of worldviews and the connection toreform efforts often found in formal settings.Echoing research that suggests the crucial role ofstructured tiime for teachers to meet (Louis et al.,1996), the experiences of teachers in StadeleElementary suggest that the policy world can en-courage productive collaboration by providinrgsufficient funding for teachers to work with oneanother without too many demands for whatteachers are to do during this time (Stokes, 1997).Stadele Elementary supported teachers workingwith one another by using grant ftnds that werefinite and relatively short term. But even withthese additional funds, pressures to complete as-signments often cut short conversations just asthey began to move to a deeper level. Increasedexpectations that teachers will work together on'joint work" (Little, 1990) without resources andtime to support that work in a meaningful waymay lead to the kin(d of out-facing conversationsthat prevent teachers from engaging deeply withideas or questioning preexisting ways of doingthings.
Third, beyond resources for time, the experi-ences of Stadele Elementary suggest that reformleaders should structure collaboration around
163
Coburni
authentic activities that have clear connections tothe classroomn. Fourth, teachers in formal settinos
otteni have diverse worldviews and ways of teach-ing. At Stadele Elementary, this pedagogical di-versity led to cotflict or avoidance of differencein ways that contributed to out-facing conversa-tions. Teachers need structures and support thathelp them engage in conversation across diver-sity, to help make diverse settings opportunitiesto learn fromii one antother and push thinkingrather than places to disengage and avoid con-
flict (Achinstein, 1998)."
Finally, policy can find ways to provide greateraccess to knowledge resources. Too often. policyand professional development fail to provide suf-ficient resources to help teachers understanid new
approaches or mEaterials in sufficient depth to be
able to mnake the kinds of principled professionaljudgments necessary to bring themn into theirclassrooms (Cohen & Hill, 2000; Spillane & Jen-nings, 1997). Documents, compendiLuns of as-
sessments a-nd model lessons, textbook series
with scant professional development, and one-shot workshops continue to be all-too-commnonways that policy attempts to iniluence teacher
practice (Cohen & Hill, 2000; Little, 1993). Ac-
cess to external knowledge resources may be es-
pecially important when policy is promotinginstructional approaches that are likely to be
unfamiliar to teachers. At Stadele Elemeentary,external knowledge resources became increas-ingly influential when teachers did not feel
they had knowledge resources available in their
proximal communities. Furthermore, Spilaltne
(1999) suggests that connections to individuals
with deep knowledge of reform. practice maybe a key attribute of collective settings that en-courage teachers to mnove beyoind incrementalchanges in practice.
But beyond providing sufficient support and
depth to professional development accompany-
ing miessages about readiing comiing fromn theenvironmlent, teachers also need access to greater
knowledge resources at the school site on an on-going basis. This calls for finding ways that pol-
icy can support the continuous developtmtent of
collective knowledge resources at the school so
that teachers not only have places to turn at the
school site when they find thetnselves with ques-
tioIIs about their practice, but also can continue toexpanid and extend their worldviews and repertoireof approaches to reading instruction over timre.
164
Notes
I wish to thank Milbrey McLaughlin. Joan 'Falbert,Bettv Achinstein, Laura Stokes, Nathani MacBrien,Matthew Janger, and Jennifer Borinan for commentson an earlier draft of this paper. I also want to thank theeditors of EIPA and three anonymous reviewers fortheir especially thoughtful comments. Data collectionwas supported by the Center for Research on the Coni-text of Teaching at Stanford Utniversity. Data analysisand writing were supported by a dissertation fellow-ship from the Spencer Foutndation.
I Vaughan (I996) defines worldview in the follow-ing way: "Each person-the butcher, the parent, thechild-occupies a different position in the world,which leads to a unique set of experiences, assump-tions, and expectations about the situations and objectsshe or he encounters. From integrated sets of assump-tion s, expectations, ancd experience, individuals con-struct a worldview. or frame of reference, that shapestheir interpretations of objects and experiences. Every-thing is perceived, chosen, or rejected on the basisof this framework" (pp. 62-63). For the purposes of thisstudy, "worldview" should be distinguished from themore conventional notion of "belief' in several ways.First, "worldview" moves away from any static asso-ciations the term "belief' miight carry. highlightingthe ways individuals' histories and particular sets ofexperiences shape the ways of thinking or cognitiveframes they use in interpreting events and information.Furthernore, these events and information, in turn,shape worldview over time. Second, the concept of"worldview" acknowledges the influence of profes-sional norms in the institutional environment andshared understandings that are deeply situated in theparticular organization or workgroup.
2 Stadele Elementary is a pseudonym.3 The student body at Stadele Elementary School is
very diverse: 55% are, Asian (of whom the majority areChinese), 28% are Latino, 13% are African American,and 3% are white. Sixty-seven percent of all studentsqualify for free or reduced-price lunch, and 48% areclassified as English language learners. The teachingfaculty is also racially diverse. Of 33 classroom teach-ers, 48% are white, 33% are Asian (Chinese and Fil-ipino), 12% are African American, and 6% a re lJatino.All five resource teachers are white and the principalis Asian American.
I Reading Recovery is a pull-out tutoring programiin which trained Reading Recovery teachers workone-on-one with the first-grade students who are as-sessed as the lowest readers.
I There are trade-offs involved in such in-depth ob-servations. On the one hand, this kind of observationprovided a depth of understanding of teachers' world-views and practices as well as access to teachers' in-formal relationships with their colleagues in a way notpossible fromD one or two short observations. (In the
Collective Sensemakling about Reading
other hand, investing heavily in observing fewerteachers meant that I was unable to observe all teach-ers in the early grades. But although I did not observeall teachers, I gained much informatiosn about theirclassroom practices through interviews and, peerhapsmore imiiportant, through extensive meeting observa-tions. Observations of teachers engaged in joint workrelated to classroom instruction proved an especiallyfruitful source of information about teachers' world-views and practices. Activities such as developing andscoring studlent assessments often revealed teachers'assumptions about appropriate reading instruction andthe nature of student learninig, as well as the details oftheir practice, in a way that interviews did not. Myconfidence about the information gained about world-views and practices in this manner for teachers whomI did not observe was bolstered by the high level ofcorrespondenice between classroomr observations andwhat teachers said in meetings among the teachers 1did observe.
6Reconstniction of state messages happened insome cases in spite of aggressive efforts by the state tocontrol the messages about reading. For example, thestate attempted to control the content of professionaldevelopment funded with state monies. Under AB 1086,professional development providers had to submit pro-fessional development plans to the state to be approvedin order for districts to contract with them. However,districts could bypass this approval process by creatingtheir own professional development. Stadele Elemen-lary's district took this other route, developing profes-sional developmenit for all K-3 teachers in a way thatreinterpreted the state's heavily skills-based emphasis.
I For example, during the year of the study, teach-ers at Stadele Elementary School received profes-sional development for the new reading series that wasadopted by the state in 1996 using criteria developedby the state in 1994. While it could be argued that theniew reading series represented a "balanced approach"in its broadest sense, key parts of the reading series anidthe 1994 criteria contradict the rendition of the bal-anced approach in current state policy in significantways. These differences can be illustrated by compar-ing the 1994 criteria with criteria adopted in 1999 forthe 2002 adoption (which are miore representative ofcurrent state policy). These two documents differ intheir view of appropriate reading materials to use (pre-dictable texts versus decodable text); whether skillsare to be taught in context or in isolation; whetherskills should be presented in a prescribed sequence oron an as-needed basis; in the emnphasis on a model ofreading emphasizing three cuteing systenms versus amodel of reading emphasizing the graphophonic sys-tem; and in promoting a model of instruction in whichteachers are facilitators guiding student constructionof knowledge versus a model of reading in whichteachers provide direct instruction and students then
practice skills by reading (Cuirriculuni Commission.1999; Curriculun l)evelopment and SupplemnentalMaterials Comnission, 1994). In spite of the ways inwhich the new reading series contradicted state policyin 1998, textbooks were one of the key routes bywhich teachers in this district were connected withstate policy messages, illustrating the time lag.
I Teachers at this school had worked in foiral grade-level groups during previous reform initiatives and hadbeen involved in the reform project since the previousyear (with few changes in personnel). Some of the in-fonnal groups had a much longer history of workingtogether-in one case stretching at least 15 years.Thus, when new practices, ideas, or approaches en-tered into teachers' conversations with one another,they entered conversations that had a history, withnonns of interaction and shared understandings thathad developed over time.
9 Hill (1999) reports a similar pattern in her studyof a district committee's efforts to write district math-ematics standards that were aligned with the statestandards.
Interestingly, the differenit models of reading pres-ent in this grade level group roughly correspond to thecurrent disagreements in the policy world about whatkinds of skills and strategies teachers should use to fos-ter early decoding. Like teacher G, state. policy empha-sizes the primacy of graphophonic infornnation andcounsels teachers to focus on that information first andforemost: "Students understand basic features of read-ing. I'hey select letter patterns and know how to trans-late them into spoken language by using phonics, syl-labication, and word parts" (California Department ofFducation, 1999, p. 61). Other approaches, includingthat put forth by Reading Recovery, paint a picture of areading process by which childreii attend to a range ofcues including graphophonic (visual), semantic (mean-ing), amid syntactic (structure): "There are nmany differ-ent strategies for reading ... (1) Strategies that maintainfluency, (2) strategies that detect and correct error, and(3) strategies for problem solving new words. Eachfunction involves a network of cues provided by mean-ing, language structure or syntax, and visual infoima-tion. The important thing about these cues, however, isthat readers access and use thenf' (Fountas & Pinnell,1996, pp. 149-150, emnphasis in onginal text).
It is important to note that running records actuallyhave highly elaborated procedures for administrationand analysis-procedures that place a central impor-tance on using real books, pictures and all. However,these procedures were covered quickly in the two1-hour workshops that teachers participated in. With-out having this information at their disposal, teachersended tip creating their own procedures, which tendedto differ considerably from the approach promoted inthe professional development. I don't mean to imply,however, that greater access to training would have led
165
Cobu77?
to greater fidelity to the process. T'eachers at anotherschool in this study with considerably more training inrunning records atso altered procedures because oftime and management concerns, but did not alter themquite to the same extent as teachers in Stadele Elemen-tary School.
12 Of the 29 meetings I observed in which teacherswere intentionally grouped for activities. they weregrouped in cross-grade groups for five of them and inlgrade-levei groups for the renmaininig 24.
1 Johnson and his colleagues (1998) report a sim-ilar finding in their multilevel study ot the imple-mentation of a literature-based reading progranm infour elementary scholols in f:our districts.
14 The relationship between Reading Recoveryteachers and classroom teachers in inflrirmal networkshighlights the importance of teachers' perceptions ofeach others' woridviews and practices in developiniginformnal netvorks. As one anonymous reviewer forthis article suggested. and my interview and obser-vational data confirrn, the Reading Recovery teachersin the school had very similar ideas about the nature ofteaching reading and nearly identical approaches todoing Reading Recovery, perhaps because of the in-tensity and ongoiiig nature of their professional devel-opment and the closeness with which they workedwith one another. Yet new teachers reached out dis-proportionately to one of the three Reading Recoveryteachers based on their perception that her practice andbeliefs were different from the other two.
15 Here, the relevant unit of analysis is thie activityor assignmeent that structured conversation in teachergroups during fonnal settings. It is important not to con-flate "activity" with meeting tinie. There were some-times multiple assignments in a given meeting and, forthat mnatter, meetings without formal assignments oractivities at all. Similarly, it is important not to confuseactivity with message. Teacher groups often hrew oiimultiple messages about reading in the course of a par-ticular activity.
1I The developmenit of teacher conmmunity was a cen-tral goal of the reform organizatioii. However, StadeleElemientary placed particular emphasis on it, makingdecisions about funcding in ways that placed a centralemphasis on creating meeting time for teachers.
I I For example, grade levels were repeatedly directedto the reading series aud district materials for informa-tion anid guidance on assessment: "Look at the [readingseries] assessment kit and the assessment of the districtstandards binder and discuss runninig records as a tormof assessnieit at your grade level" (released-dav as-sigimenit sheet, 10/27/98) and "Choose a basic assess-ment piece for reading comprehension. It should besomiething you can give to your class two to three timesa year to measure reading comprehension. both recalland higher lcvel thinking skills. Suggestion: Use an as-sessmnent piece already provided by the [reading seriesI
166
assessmient kit" (released-day assignment sheet.12/1198). And, following district professional develop-ment oni assessment, "Which, if any, assessments didyou leam about that you are interested in irnplemnentingat your grade level for the whole class, just for strug-gling readers?. . .Which, if any. of the assessments doyou thinik we could choose schoolwide, K-2, 3-5?"(released-day assignment sheet 3/25/99).
'8 At the same time that the principal's construLsctionof the reading series vis-a-vis standards authorized di-verse approaches to reading instruction, it also boundedthis diversity by linkinlg it to outcomes specified by dis-trict standards-outcomes that, in turn, were assessedby standardized test scores. During the year of thestUsdy, the school was grappling with district standardsfor the first timne. In a school with a strong tradition ofteacher autonomy and multiple and diverse concep-tions of what "good" reading itistruction entails, the no-tion of standards was controversial. As teachers in theschool discussed the nature and implications of stan-dards for their classroonms, a small number of teachersvoiced the opinion that gearing instruction to standardswas a fornm of standardization that encroached on theirautonomy. Others, including most of the teachers inthe early grades, supported the neotion of standards asa mecharnism to create clear grade-level expectationsand align instruction across grade levels. Over thecourse of the year. in conversation and interaction withthe standards, the teachers and the school leadershipworked to negotiate the perceived tensions betweenstandards and professional autonomy-a tension thathas also been the topic of debate in the policy world(see i)arling-llammond, 1997, for a review). Whilethey had by no means resolved this tension by the endof the school year, through this process of interactionthe faculty had begun to construct a conceptioti ofstandards as providing goals or loose guidelines thatcan be met through many differenit approaches toteaching. In this sense, they redefined professional au-tonomy as the autonomy to make decisions abotut theinieans to get to outconies defined by standards. Sig-niticantly, this stance towards standards was, in turni,mnade possible by a district that saw it as the principal'srole to make decisions about textbook use and hlow theschool actualizecd the standards.
I' Professional norms of autonomy and privacy inteaching leave maniy teachers ill prepared when col-laboration surfaces differences and conflict. Achini-stein (1 998) argues that schools must establish explicitnormrs for interaction that value and indeed embracedifferenices in opinion, and use inquiry and decision-making structures that provide processes tor teachersto tailk about, explore, work with, and rnediate theirdiffering worldviews and practices. One such structurefor inquiry-and there are many-is the Tuning Pro-tocol, developed and used by the Coalition of Essen-tial Schools to guide teacher inquiry and collaboration(Allen, 1]995).
Collective Senseinaking about Reading
APPENDIX
TABLE AlDistribution of Meeting Observations
Hours of observation
Formnal settingsFirst-grade meetings 14.5Second-grade meetings 16.0Other grade-level meetings 13.5Full faculty meetings 43.5Leadership team meetings 13.5In-school professional development 15.0District professional developmenit 3.0)External professional development with leadership team 11.5Total 130.5
Inforimal SettingsInfonnal meetings Ongoing
TABLE A2Distribution of Interviews and Classroon Observations of Teachers
Hours ofYears of Leadership Number of classroom
Individual Teaching level experience Sex/Race team'? interviews observation
A 2nd grade 29 F/Asiani N I OB 2nd grade 7 M/white N 2 0C 2nd grade 4 F/Asian N 12 33D 2nd grade 5 F/white Y 4 5E 2nd grade 3 F/Asian N 0 0F 1 st grade 36 F/white N 3 5G 1st grade 34 F/white N 10 23H Ist grade 3 F/African N 0 0
AmericanI 1 st grade I F/white N 7 253 I st grade 3 F/Asian N 1 0K 1 st grade 2 F/Latina N 1 0M K 33 F/white Y 2 0R Reading 3 1 F/white Y 2 5
RecoveryS Reading 14 F/white Y 2 5
RecoveryT Reading 17 F/white N 1 5
RecoveryU 4th grade 34 F/white N 2 0V 4th grade 34 F/white N 2 0W Reform 7 F/white Y 5 N/A
coordinatorX Principal 5 F/Asian Y 3 N/A
167
Activity
Coburn
References
Achinstein, B. (1998). The Ties that Bintd. tJnpub-lished doctoral dissertation, Stanford ULniversity,Stanford, CA.
Allen, D. (1995). The Tuninig Protocol: A Process forReflection. Providence, RI: T'he Coalition of Essen-tial Schools.
Barley, S. R. (1986). Technology as an Occasioni forStructuring: Evidence from Observationis of CTScanners and the Social Order of Radiology Dc-partments. Administrative Science Quarterly, 31.78-108.
Barley, S. R. (1990). Images of Imaging: Notes onDoing Longitudinal Field Work. OrganizationalScience, 1, 220-247.
California I)epartment of Education (1995). EveryChild a Reader. Sacramento: Author.
California Department of Education (1999). Reauding!Language Arts l ramewo-ork br California PublicSchools: Kindergarten thr-ough Grade Twelve.Sacramento: Autlor.
Californria State Board of Education (1999). Guide tothe California Reading Initiative 1996-1999. Sacra-mento: California Reading Initiative Center.
Coburn. Cynthia E. (2001). Makiig Sense of Reading:Iogics of Reading in the Institutionial Enl vironmentanti the Classroom77. Unpublished dloctoral disserta-tion, Stanford University, Stanford, CA.
Cohen, D., & Ball, D. L. (1990). Relations BetweenPolicy and Practice: A Commentary. EducationalEvaluation anid Polic yAnalysis, 12, 331-338.
Cohen, I). K., & Hill, H. C. (2t000). lnstructional Pol-icy and Classroom Performiance: 'Tlhe MathematicsReforrm in California. Teachers College Recorat,102, 294-343.
Cuban, L. (1988). Constancy and Change in Schools(1880s to the Present), In P. Jackson (Ed.), C'on-
tributing to Eduication-ial Change. Berkeley, CA:MeCutchatn.
Curriculumn Commission (1999). Criteria for 2002 Lan-guage Arts Adoption. Available online: http:/Avvww.c(le. ca.gov,/cilbratch/eltdiv/2002lacriteria.hrlm.
Curriculum Development and Supplemental MaterialsCommission (1994). K-8 Language Arts Instruc-tional Resources Evaluationi Form. Availableonline: htt)://lvww.cdce.ca.gov/cilbrtanhlengiang/
ela_adopt.htnt.Darling-Hammond. L.. (1997). T''lhe Right to Learn:
A Blueprint for Creating Schools that Work. SanFrancisco: Jossev-Bass.
EEPA. (1990). Educational Evaluation and PolicyAnalysis, 12, 233-353.
Fisenhart, M. A., & Howe, K. R. (1992). Validity inEdtucational Research. In M. D. LeCompte, W. L.Millroy. and J. Preissle (Eds.), The Ifeandbook oq
168
Qualitative Research in E'ducation. San Diego, CA:Academic Press.
Fountas, I. C., & Pinniell. G. S. (1 996). Guided Reading:Good First Teaching for All Ci'hildrenz. Portsnouth,NH: Heinemann.
Fuhrman, S. (Ed.). (1993). Designing Coherent Eda-cation Policy: Improving the Systrem. San Francisco:Jossev-Bass.
Goodlad, J. 1. (1 984). A Place Called School: Prospectsfor the Future. Mahwah, NJ: Lawrence ErlbaumAssociates.
Hargreaves, A. (1994). Changing Teachers, ChangingTimes: Teachers' Work and Culture in the Post-modern1 Age. Lonidon: Cassell.
Hartley, J. F. (1994). Case Studies in OrganizationalResearch. In C. Cassell & G. Symon (Eds.), Qualita-tive Metrhods in Organizational Research: A Practi-cal. Guide (pp. 208-229). Thousand Oaks, CA: Sage.
Hiill H. (C. i1999). Policy is not enough: Language andthe interpretation of state standards. UJnpublishedmanuscript.
Isabella, 1.. A. (1990). Evolving Interpretations as aChange tUnfolds: How Managers Construe KeyOrganizational Events. Academy qf ManagemientJournal, 33, 7-41.
jennings, N. E. (1996). Interpreting Policy in RealClassrooms: Case Studies of State Retbrmt andTeacher Practice. New York: Teachers CollegePress.
Johnson, P., Allington, R. L., G(uice, S., & Brooks,G. W. (1998). Small Change: A Multilevel Studyof the Itmplementation of Literature-Based Instruc-tion. Peabody Journal oj'Education. 73(3 & 4),81-103.
Lin, A. C. (2000). Reform in the Making: The Inple-mnentation otfSocial Policy in Prison. Princeton, NJ:Princeton University Press.
Little, J. W. (1982). Norms of Collegiality and Ex-perimentation: Workplace Conditions of SchoolSuccess. Amierican Educational Research Jousrnal,19, 325-340.
Little, J. W. (1990). The Persistence of Privacy:AuLtonomy and initiative in Teachers' ProfessionalRelations. Teachers College Recorl, 91, 509-536.
Little, J. W. (1993). Teachers' Professional Devel-opment in a Climate of Educational Reform. Edu-cational Evaluation and Policy Analysis, 15(2),129-151.
LIittLe, J. W., & McLaughlin, M. W. (Eds.). (1993).Teachers'Work: Individuals. Colleagues, and Con-texts. New York: Teachers College Press.
ILofland. J. & Lofland, L. l. (1995). Analyzing SocialSettings: A Guide to Qualitative Observation andAnalysis. (4th ed.) Belmont, CA: Wadsworth.
Lortie, D. C. (1975). Schoolteacher: A SociologicalStudy. Chicago: University of Chicago Press.
Collective Sensemaking about Readinig
Louis, K. S., & Marks, H. M. (1998). Does ProfessionalCormmunity Affect the Classroom? Teachers' Workand Student Experiences in Restructuring Schools.Amaericai Joournal of .education, 106, 532-575.
Louis, K. S., Marks, H M., & Krutse. S. (1996). Teach-ers' Professional Community in RestructuringSchools. American Educational Research Jousrrnal,33, 757-798.
Malen, B., & Ogawa, R. T. (1988). Professional-Patron Influence on Site-Based Governance Cotuni-cils: A Confouniding Case Study. EducationalEvaluation axnd Policy Analysis, 10, 251-270.
McLaughlin, M. (1991 a). Learning from Experience:Lessons from Policy tmplementation. In A. R.Odden (Ed.), Education Policy Implementation(pp. 185-196). Albanv, NY: State University ofNew York Press.
McLaughlin, M. (1991b). The Rand Change AgentStudy: Ten Years Later. In A. R. Odden (Ed.),E'ducation Policy Implementation (pp. 143-156).Albaniy, NY: State University of New York Press.
McLaughlin, M. (1993). What Matters Most inTeachers' Workplace Context? ln J. W. Little &:M.McLaughlin (Eds.), Teachers' Work: Individuals,Colleagues, and Contexts, (pl. 79-103). NewYork: Teachers College Press.
McI.aughlin, M. W., & Talbert, J. E. (in press). Pro-fessional C ommunities and the W7ork of figh School'Teaching. Chicago: University of Chicago Press.
McQuillan, Jeff l1998). The Literac'y Crisis: FalseClaims, Real Solutions. Portsmiiouth, NH: Heine-nmann.
Meyer, J., & Rowan, B. (1 977). Institutionalized Orga-nizations: Formal Structure as Myth and Ceremony.American .Journal of Sociology, 83, 340-363.
Meyer, J., & Rowan, B. (1978). The Stnicture ofEducational Organizations. In M. W. Meyer (Ed.),Environments and Organizations. San Francisco:Jossey-Bass.
Miles, M, B., & Huberman, A. M. (1994). QualitativeData Analysis: Ani Expanided Sourcebook. (2nd ed.).Thousand Oaks, CA: Sage.
Newtiann, F. M., & Associates. (1996). AuthenticAchievement: Restructuring Schools for IntellectualQuality. San Francisco: Jossey-Bass.
Odden, A. (Ed.).(199 1). Education Policy Implemen-tatioti. Albany, NY: State UJniversity of New YorkPress.
Porac, 1. F., Thomas, H., & Baden-Fuller, C. (1989).Competitive Groups as Cognitive Commnunities: TheCase of Scottish Knitwear Manufaicturers. Jourcnal ofManagement Studies, 26(4), 397-416.
Pressman, J. 1.., & Wildavsky, A. (I 984). Implementa-tion. Berkeley, CA: tUniversity of California Press.
Rowan, B. (1982)1 Organizational Structure and theInstitutional Environment: The Case of Public
Schools. Administrative Science Quarterlv, 27,196-198.
Sarason, S. (1971). 7he Culture of School and theProblem of Change. Boston: Allyn and Bacon.
Scott. W. R. (1995). Institutions and Organizations.I'housand Oaks, CA: Sage.
Scott, W. R., Mendel, P., & Pollack, S. H1996, March).Studying "Eveiything Else ": Assessing thEe Environ -ments of an Organiizational Field. Paper presented atthe Conference on Institutional Arnalysis, Universityof Arizona, Tucson, AZ.
Siskin, L. S. (1994). Realms of Knowledge: AcademnicDepar-t.ments in Secotndlary Schools. Washington.DC: Falmer Press.
Smith, M. S., & O'Day, J. (1991). Systemic schoolreform. In S. Fuhrnman & B. Malen (Eds.), The Pol-itics of Curriculumt andu Testing (pp. 233-267).New York: Faimer Press.
Snow, D. A., Rochford, E. B., Worden, S. K, &Benford, R. D. (1986'. Frame Alignment Processes,Micromobilization, and Movement Participation.American Sociological Review, 51, 464-481.
Spillane, J. P. ( 1996). School Districts Matter: LocalEducational Authorities and State InstructionalPolicy. Educational Policy, 10(1), 63-87.
Spillane, J. P. (1998). State Policy and the Non-Monolithic Nature of the Local School District:Organizational and Professional Considerations.Americant Educational Research Jouirnal, 35(1),33-63.
Spillane, J. P. 1 999). External Reform Initiatives aiidTeachers' Efforts to Reconstruct their Practice: TheMediating Role of Teachers' Zones of Enactment.Journal of Curriculum Studies, 31(2).
Spillane, J. P. (2000). Cognition and Policy linple-mentation: District Policymakers and the Reform ofMathematics Education. Cogniition and Instruction,18, 141--179.
Spillane, J. P., & Jennings, N. E. (1997). Alignedlnstructional Policy and Ambitious Pedagogy:Exploring Instructional Reform from the ClassroomPerspective. Teachers College Recordl 98, 439-481.
Spradley, J. P. (1979). The Ethnographic Interview.Fort Worth, TX: liarcourt Brace Jovanovich.
Stokes, L.. (1997). Short-term Policy Support for Long-term School Change: A Dilemma for Reform-mindedPractitioners. journal of Education Policy, 12,371 -384.
Strauss, A., & Corhin, J. (1990). Basics of QualitativeResearch: Grounded 7'heory Procedures and Tech-niques. Newbury Park, CA: Sage.
Talbert, J., & McLaughlin, M. (1994). Teacher Pro-fessionalisni in Local School Contexts. American.Journal of Edlucation, 102, 123-15d3
Tyack, D.. & Cuban, 1,. (1995). Tinkering TowardUtopia: A C'en tury of Public School Reform. Cam-bridge, MA: hlarvard University Press.
169
Cobarn
Vaughan, D. (1996) 7he ChallengerLcLunch Decision:Risky Technology. Cult{re, and Deviance at NASA.
Chicago: LJniversity of Chicago PressWeatherley, R., & Iipsky, M. (1977) Street-Level
Bureaucrats and Institutional Innovation: Imple-menting Special-Education Reform. Iarlcvard Edit-
cational Review, 47, 171-197.Weick, K. E. (1995). Sensemnaking int Organizations.
Thousand Oaks, CA: Sage.Yaniow, D. (1996). HIow Does a Policy Meani? lnter-
pretiug Policy and Organizational Actiorns. Wash-ington, DC: Georgetown University Press
Yin, R. K. (1984). Ca.se Study Research: Design, andMethods. Beverly Hills: Sage.
Author
CYNTHIA E. COBURN is an assistant professor inthe IDepartmient of Administrative and Policy Studiesat the Uniiversity of Pittsburgh and a research scientistat the Learning Research Development Center, 3939O'Hai Street, Utniversity of Pittsburgh, Pittsburgh, PA15260; e-mail: cecoburn7@pitt cdu. She specializes inthe relationship between policy and practice, urbanschool reformn, and qualitative research methcods.
Mainuscript received August 10, 2000Revision received January 23, 2001
Accepted February 28, 2001
170
COPYRIGHT INFORMATION
TITLE: Collective sensemaking about reading: how teachersmediate reading policy in their professional communities
SOURCE: Educational Evaluation & Policy Analysis 23 no2 Summ2001
WN: 0119603465003
The magazine publisher is the copyright holder of this article and itis reproduced with permission. Further reproduction of this article inviolation of the copyright is prohibited. To contact the publisher:http://www.aera.net/.
Copyright 1982-2001 The H.W. Wilson Company. All rights reserved.