Dressman_Research_Evidence - Mark Dressman Uniserst'flr...

Info iconThis preview shows pages 1–28. Sign up to view the full content.

View Full Document Right Arrow Icon
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 2
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 4
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 6
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 8
Background image of page 9

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 10
Background image of page 11

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 12
Background image of page 13

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 14
Background image of page 15

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 16
Background image of page 17

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 18
Background image of page 19

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 20
Background image of page 21

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 22
Background image of page 23

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 24
Background image of page 25

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 26
Background image of page 27

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Background image of page 28
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Mark Dressman Uniserst'flr ofliit'rtots, UfbdM—Cbflmpflfgn, USA Reading Research Quarterly Vol. 3.4, No. 5 Julyfnugusta’S-eptmher I999 @1999 International Erasing MUM (pp. 255—355} On the use and misuse of research evidence: Decoding two states’ reading initiatives thin the fields of school literacyr research and policy making. disagreements about the relationship between spoken and writ— ten language in the teaching and learning of beginning reading have seldom been as heated or as franght with issues of power and control as in the 1990s. As Pearson (199?) and others have noted. even in the 1960s at the height of arguments over reading instruction and in the midst of enormous social and political unrest within the United States, a spirit of collaboration rather than contention usually characterized discourse within the reading research community itself. In his comments. which. accompanied Reading Research Quarasflfs recent revisitation of the First Grade Studies (the massive, feder- ally funded comparative—methods project meant to re- solve the issue of best practice in reading once and for all [Bond St Dykstra, 1965'31991'; Pearson, 1997'; Readcnce St: Barone, 199Tl'l, Pearson implied that this collaborative spirit was the consequence of a shared sense of the im- portance of the undertaking by The researchers involved at all that study's sites. But what Pearson does not acknowledge is that three assumptions were implicit in nearlyr all the ap- proaches compared within the First Grade Smdies: (a) that reading consists of two processes, word recognition and comprehension; (b) that word recognition is the req- uisite of comprehension: and {c} that the orthographic complexity of written English appears to be a major stumbling block for beginning readers. In contrast, in the 199135 it is possible to argue with integrity and force that literacy acquisition can more aptlyr be described as a de- 255 velopmental, rather man as an associative, process (Dyson, 1993; Teale 5': Sulrhy, 1935]; that the drive to comprehend is the impetus behind word recognition (Goodman, 1996}; and that the communicative practices ofa social group have as much or more to do with corn- plitating literacy instruction and acquisition as does or- thography (Au, 1996; BarreraI 1992; Cazden, 19%: HeathI 1935; l'tliiilltinson. 1998). Each of these alternatives depicts beginning readers as complex social and cultural actors and reads the highly qualified facts derived from the psychologically oriented comparative-methods stud- ies of the 1960s as sure signs that, when studied in their fiJll contexts, the social, cultural, and political Factors in students‘ lives and school careers may be as crucial in determining how and wlmther they acquire literacy as any instructional approach flianolomé. 1994}. As a conseouence, in the 19905 what appears to be indisputath objective scientific knowledge about early literacy to some appears to others to he a set of discrete facts that have been broadly interpreted to produce poli- cies and literacy curriculums that are as much the prod— uct of their makers‘ cultural politics and nonnative assumptions about social reality as they are the product of a dispassionate use of the scientific method. To take the First Grade Studies as a prime example of this, it has long been an article of faith among reading researchers that the LLB. federal government's interest in reading was first motivated by the cold war rhetoric of Rudolf Flesch {1955, 1965) and the geopolitical implications of Sputnik. But Arlene Willis and 1v'ioliet Harris {199?}. two African American researchers who grew up and learned to read t‘tEIE-ETH'IHLZ [Er THIS STUDY investigates the claims oi scientific chjecdyiry that sup- port tenet" changes in policies toward early literacy instill-Mimi in me states of Texas and dah‘tornia. According to recent reading initia- tjees in both stats, conclusive findings in literacy research. particu- larly in the area of phonemic-“phonological awareness [PM], nou-r rnandate state policies of explicit, systematic inst in phonics and the use oi phonetically regular terms in early grades .‘t [lame- wori’. suggested by Haberrnas's {193?} Theory ty'ti'ont maximum nation was used to assess the validity of this Claim. though the analysis of 1D major shadies of PM. 2 sentinel reviews of early liter- aey research, and 2 policy documents Close analysis of these tests generally supported the obiectit-e claims for PM research but chal- lenged the poor performance of minority artd lw-seeweronomic ESTE ESTUDIO imestiga los aspuesltei de obit-Livth cientilica que Fundarrtetttan caruhios retientesat [as puliicassohe la slob-cum 'tnicial en los eateries the Texas y California De acueldo tur- Ins in.'i- eiatit'aa recientes adoptadas en anisos estados. exists evidencia con- elusn'ea en la investigacion sohre alfahetizaclon, partlcutartrtente en cl area dr: comicncia Foncmitaflortulogita. one denunth politttas es- tataJes de enset’tanza fortita explicita y sistenritica y el uso de testes toasétimrnente regulates en ios grades iniciales. Se utillao el ntarco sugeridopmla Lemiadehacodncmnmucamodeflabennasilfi} para what it 1ralidci: de est: supueslo. a pallirnlcl militia de m E:- tenaos estudios decortciencia fonoldgica. 2 itnpmantes revisioues so- bre la intestigacidn en alfabetizacitin initial y dos dot'ttnterttos die politics educat'n'a. En térrrtinos generalea el analist's detaliado de estos leatus apoya Ia objeaiyidatl de los supuestos the la inuesu'gacidn en tutttiertcia lorruldgica. peru lamhiht present: un desafio :1 mal de- DIIEE S'I'IJDIE urttersudtt die r'u'tsprliche aufwissenscltafiiicl'tcrfibiek- tit-ital, weld-re hit-elicit durchgefijhrte Andemngen bei "r'erordttungen Ella-er [riihe Lose- und Schreibanwcimngen in den Slaaben Texas tmd Kalifntnien LimE'I‘Sl'CIIIE'fl. .itufgruod hrtzlicher Leseinit'atisen in beiden Staa‘ten1 Mugflndflt Wm inder Potsdnt uher das lesert urtd Schrejben, oesonders au!‘ tiers Geociet des phonemtschetupho- nologischen Aufl'assungsbewufitseins [PPM machen ste es ietzt ntr Pflicht, nae die staatljcit yerotdnetett, ausfhhrltchen Richtljniert urtcl snweteungen fil‘trd‘il: thrtenu’lt tlnltalm- Lattte unddetu Gehmueh all- gemein-phonetiseher Test: in den Eirtsthtllttngss'htfen befolgt 1Mer- dett. Ein liahntenkortzept nath i«'orslzhligest you Habertnas’ [195?) MmrfirMmmMmflm mode angewaasti. untdje grudge Bewein‘i'thmngdiestsansprucltsrubeserten, dureitdtefimh-setm IEI Hauptshtdnm fiber EPA, 2 hileé‘ItiEEn Iserrt'trtalt-n] Kt't'tiken iibcr iniherltettntlidae Lesa- und Schreibrforsehung. suwie 2 Dulcuntenta- tiotnert air 1tierfahrertsstreise. Eine nahere Analyse dieserTetrte unher- srutzte irn altgerneirten cite Uttjeinisdtit der aussagen ntgunaten tier PPJL Forschung. machten jedoch Einwentlungert in bee-up aul das Dntbemnndmfimequambeuflemmflecofinghnnsflfles'madfltghflrm status [SE51 populations on tests at PPJL on are grounds that differ ences in the norms of those pupulatiuns' phonological systems may bcias ten. raultst Addih'ottally. obierahrely referetvzod claims made in two major research resdews that. me pmr performance of nonrrtain- stteatn students on m HP“. and reading acJtietement are linked totl'te social orgertetic inferiolitycftheatudems' Eamt'liesandcntu- rnurtida appeared to begrounded in tuttcstod and urracln'torwlcdgod numeric: assumptions about the home litres and genetic baclh grounds of children who struggle to learn to read. A concluding. anatyn's ot‘ the two states' ElllTifiJLiI policy Hetements found them to be highly selective in their use of researtit evidence and, from a Habettnasian perspective. more strategic than mmuticadte in their orientation and intent. Asmarddbuertymtuso rte [a erdriencirt ciaufificmdtrdiiaiardehrs tittidaflurtspm'a in metro- puestas err dos eateries semero de. los ninos pertertocietttes a rninorias y a niyeles socio- econdtuicos betsz en las pruehas de conciencia formldgica. Ello se bass en que las dilerenc-ias en las rtorrnas de los sistetnaa Fonologi- ous the cats pohlaeiotaes podr‘tan sesga: los resultados de las pree- bas. Adernas. los supuestos biett hand-Intent:st dc tlos imponantes re't'is'tona de la lnyestigau'dnaoercade one el real deserrtper‘to de los estudiantes tle pohiachants sanctions: en pruebes dc mnciencia fortuldgitaylotturaestauinculadoa la inferioridadsocialyrogcnfli- cadelastanu‘liasyctanurudades. pareoenestartasadosen omoep- eiortes no evaluaths obietjyamente solute Ins endless de mi: y t:- racletitt'eas genetieas ale nines que iuehan per aprender a beer. Para cutvcluir. un utilisis de los fundamenth de la poiltica curricular de arnttcts estados reyela qtte sun ahatnente seJecdyos en el uso tie ]a euldertcia cienti'ica y, descle una perspectnra llabermaeiana, mas tut- qut: mounds-us cn SJ. ot'iertlacifi't y propérsitu. filter Schmuck trad munch neat Fombuugs‘ergehtisaen: Err-tamed: truer Latte-me a‘ttleier m sthlochte hbachneiden bei Hinderheiten und soaiabchwachen iSES] Ter'len tier Beuuilremng in den PPA. Tests aufgrund detect-t. dais shwfidtmtgen yonder Numbeijmen Bet-oilserungaadtiehten rrtit dif- ferenziereriden phonologisdten Systentett die Testerge'hnisse un- Hinatitommendiealsobjeiaiy angemomenen Beltauptungert, die inzwei Haupd'u'sclamgdserflaen erhobecn muden. da‘s die schnradtett Ie'stungert yon Sunflmten der Minderheit in PPA Prdl'unpen und in Lesoleistungen rnil tier smialen undfoder Mam Unleflegenheit der Familien ttnd IGermain- sehaftett diesertihtdenten in ‘t'crbirtdung gebratht wurdett E's hat den AristitellL rhfidies in nicltt iibetprfll'baren urtd nidtttestatigten nomi- mmuhroenflberdie mm [Ebenshodingut'lgtnmfl den sbnannuungshjntergnmd [Em Etude: WI! set. weld-re beim lat-men urI:| 13:11 111i lingm. Eine sd'tblissifi: Analyse tlerhtlssagm mden Leltrplatt-Rizl'rtlinien beider marten etgah. daEsie hochgradig seleittit'itt determination; tun Forsclttmmbeweisen tuttiI aus der Haberrnas Perspelttiye, mehr strategisch als ltomrmtntlrady in fltrer Raining urtd Absicht sind. 259 mfififlmmfifi:zcmfiuaueu—F+>§filmfi7ndfine fifibr :mflfiiic ¥$fi1fi1k 511'}? as JLLTHT fifififléfltflflfifimflaiéfifitflf fiflfiflfl#fi$flfiflhatwfiifiuflw ?1F?£WflflfifltlkflMLfflcEF 91?hfl£flflfim#?2bfi%Efi¢tfl ficfliéfltfiflaetwfifimcfltLfi ctngschssimunwsuu?+ Vyfiimfifufaeuxsk.fi$$ufl wTTer§R&%fi#flflfimflfiiLs fiflmiszbefififimfeewfimmfifi H.H&I$EMT$WEflfiEfiiRU#E Eiififilflfififwfififlfiflifififli aTIdHEflTw&tLTM$o:fl$Em fififiifififitht.mmmmsflflfi m:s::¢s47-7¢92yfifimwfla EE?TH¥EM?$10M$¥&W£~MM flfimfiaiéEMTazamflflmcfifil flx2flfl§Lwfifi7nffihfififiEflmn :nesfififlfifiLtE£.Hmmfiflu$ dwtfifimifiuvwcufiLrfifieee fiwms74xus4—fifififlfiififimm hesmchat.eau2¢m$&flflifi Tusfififléfifiifififiarc$fifim$ fifl‘ FREQ-1"}; FWJ'F-F’Ar ##‘EEL—Iflfi Efiamuifimisafiflcsel;isu mfififimfliitulflEMEEEMELf watifiwa$#.%nflflaHE#flfi tfiflfiflLTwa¥mtEmifiifi$flfi mfififlawrmfli$flfifkwt—Emfi mufidwfwalfiuflbnascmzflm HmwueJaAH#EfifiLtfifiaLcs fhaflfifififlifimmuflfiLTMfiLf BU‘HmflmmfiMEfififlét.fWfifi TEJPEE‘EIEEHT: 3. 1:!151 7.1: It“? .1: l') fiLaflfififhatfiiés Apmmmm5mflmhumufihflflnflmflwflwmflfimflpmummMMHmkmmmflfiwx Enumwmmm EEITE EJLDE s'Lnfitrcssc 11.11 prfircnfims d'ohicctlvs'fl: scicnlifiquc qui appuien! les changements racents de l: polithue coneemam I'enselgnement de I'emrEE clans I'ecrlt clans les Ems du Team 131 de CahFIJmic. Sekm é: rEfiEnlllfi in:[iali.w:a en lemure prisu Clans les deux Ems. Ics cmclusimu finaJcs dues mlmchcs en “muscle. en paniculier clans le domaine de la conscience phmemlquex' phonologique [EFF], mmmandem maintenan: Ime polithue offi- cielle d'EnSEIEnen‘IEnI explicne. Systémathue de La phflnfliqut e1 I'cmplrxi nit 1:1th [Egulicrs su: I: plan phgnctiqu: clans IE5 pcfitfi classes Nous mns uliJise um: flnmrc suggtrfic per in Ibéafie die Tactic-n commuan d‘HaIJermas [193?) pmu é'lmluer h sea bd'ue de cette pIE1enILcn. en procédan: & Hanan-5e de- 1IZI [mm majeuses relatives :1 la CFP, de 2 immeanLes revues 11: question relafisfl aux Ifihtrdsca sur l'tntrfic dams l'écrit. ct d: 2 dmuments peliuques. UnE 3513135: fierrtl: dc res lexqu appuie clue maniére Eénéfilt lcs psflcntim obicctim lies recherches de CF? mais cm- Icsle Ins mus'ais résulms des enfims issus de mlmrlLés ethniques et de mflieuz sedans dfiamisés clans 135 tests de CPP du Fair que des diHéa'encB dans les nnrmes d3 systémes phmmlogiqufi :1: cc: pupulinqu pannaicnt hiaiscr IE5 résultats dcs tests. En outre. [es prerentiuns obiectl'I-remem Iéfésencées faites clans les elem: [ewe-s de queaim maiemes selcn lesquelles les mans-ass «Swims 11.11 1,35 de EFF e: the lecmre lies Eleves n'appanenan: pas au emu-am prin- C'Lpal scraltn! lies 5 me infEriOriHE sociale Won génctique lies [milks et des cumulus-auras ale res eleves apperaisscnt fondées sur des pusixlslns nonnafiwc nun testées e1 mm avErEes relatives 31 la we I'aam'liale et :1 ]'Equipemem gar-Elfin: d'eL-Lfams qui s'cl’fumun d'apprtl'drl: i lire. l'analrsl: cnfin dcs fmduncnls dc I: poliliquc dcspmgrammcs dcs dcux Ems indique qu‘ils stunt lrés sélenifs dens J'utiiisation qu‘ils font de la recherche et. clans um pesspectlve habenmsienne, cunt une curlenuticm 131 me intenutm paus strateglcpes que cmunicafimmllfl. 26-0 No states' reading initiatives 251 in the 1960s. remembered things differently. For them. the defining moment of curricular change in the Italics was not Sputnik but the Civil Rights movement, the sub- sequent desegregation of the United States’ public spaces. and in particular, its schools. and the War on Poverty that followed. Willis and Harris also pointed out that of the 15 studies chosen for comparative analysis. none mentioned race. language background. or social class as variables. They argued that this was because the only school populations included in the study were those most visible and approachable to the white. mid— dle-class reading researchers ofthe tirt‘teI that is. schools wid't whitE. middle—class students and teachers (see also Gilligan. 1982). Their revisionist perspective not only re- veals the ethnocentrism of the First Gracie Studies, it also explains ho‘tv, in a period of great social reconstruction Within the United States. the scientific endeavors of read— ing research could remain oblivious to the implications of these changes and so effectively help to preserve the dominance of white. Anglo. middle—class ways artist: cords l(Heath. 1983} in me literacy curriculums of the na- tion's primary grades well into the 1980s. it is very {erupting to take the impression from ‘Willis and Harris's reinterpretative project that they be— lieve the Ftrst Grade Studies were motivated by a grand conspiracy on the part of LS. federal officials and read— ing researchers to use the cold 1War as cover for policies designed to privilege the white middle class and ignore or deny the interests and needs of minority children. Yet they never make this accusation. nor did they ever chal- lenge the sincerity of the intentions of the researchers and officials of the 1960s instead. they offered readers a different scenario in which to set an interpretation of the First Grade Studies. In so doing, they raised a disarming— ly simple question: In hindsight. which maltes more sense in explaining the nation's anxiety about reading curriculums: connecting the Studies to an apparent exter— nal threat, or connecting them to the anticipation of imminent changes in the ethnic. linguistic. and socio- economic composition of average classrooms in nearly all of the nation‘s urban school districts? How one answers this question satisfies more than an esoteric historical curiosity. for if one decides that school desegregation is more plausibly related to the federal funding of reading research than is the specter of the cold war. then a consideration of other scenarios is also prompted. For example, what if individuals in the reading research community. acting either singly or as a group. had publicly and forcefully challenged the geopo- litical rationale and the narrowness of the First Gracie Studies' frame. and had pointed instead to the salience of internal politics and the diversity of the nation's first graders? 1Would the federal officials and researchers who produced d-re Studies then have acted to include minori- ty classrooms and teachers. and to consider tlte issues raised by them in their design and analysis? Supposing they had. then how would the Studies’ results. their con- clusions, and their influence on reading instruction have changed? Gt“, supposing they hadn't—supposing such challenges continued to be ignored even after they were madth thec would the research community at large. as well as educators at all levels. have been per- suaded to talrte the First Grade Studies seriously? These scenarios clarify the power of Willis and I-Iarris' proiect and its present signfrcance in three ways. First, lilliiilis and Harris named and were able to demon— strate the logical fallacy of claiming that if variables such as ethnicity. social class. and language background are never mentioned in a research design or report. they are irrelevant in considering the legitimacy of its findings and policy interpretation. Second. their analysis implies a method by which researchers and policy matters need not be accused of engaging in a conspiracy in order for their reasoning or habits of perception to be challenged. by demonstrating that. in instances of research and poli- cy. sincerity is irrelevant and cannot warrant the accep- tance of a finding or the legitimacy of the policies it supports. instead, they remind us mat conclusions are warranted by Whether the reasons given. when weighed against all possible reasons and scenarios. provide the most complete accounting of research findings. Third. their proth affirms that the reasons given for remarch and that policy makers provide in support of their ac— tions do matter—that reasons should not be cynically dismissed as having little bearing on outcomes, because to allow invalid reasons to go unchallenged within pub- lic discourse is to commit an effective act of complicity in the possible injustices they produce. and to surrender the possibility of rational public discourse to promote more equitable and effective educational practices. For these reasons, Willis and Hanis's approach is significant for the reconsideration it can prompt into re- cent policy—level conversations about the proper way to teach children to read and twrite. conversations support— ed by geopolitical rationales. academic discourses. and narrow habits of perception that bear distinct similarities to those of so years ago. For example, whereas in the 19605 it was the national defense to which curricular re— form ellorts were publicly reforenced and dedicated. in the 1990s the push to reform reading instmction has been supported by what Harvey Graft (1991; see also 198?. 1995} describes as the literacy myth. This is the be- lief that the rise (and fall) of nations is driven by their lit- eracy rates—-and the fear that those rates are slipping in the United States. along with the nation's political leader— ship and competitive position in die global marketplace. 262 ammo mm gunmen? Julyraiiguac’septemher 1999 stirs 1with. this rationale in mind, local and state school boards across the nation have called for the abandonment of whole language approaches and a return to traditional [that is. phonics- and skills-based} ways of teaching liter- acy lEll'lressman, McCarty, 3r. Benson, 1998}. These reading initiatives have been particularly dramatic in California. where new laws require the re-education of teachers in skills-biased approaches [California Department of Education. 1996}. and in Texas, where skills-based cur- ricular objectives are enforced by high-stakes testing pro grams (Texas Legislative Budget Boatd. 1994). Yet the economies of California and Texas are booming {Ayres 1998; Bivins. 1991'; “Winners and Losers." 199T} and high-minded connections between these states' future economic concerns and the need for curricular change seem tenuous at bestr whereas the connection hero-sen internal demographic shifts within states and their politics of school literacy seems all too obvious. Both Texas and California. for example, have absorbed large numbers of Asian and Hispanic grants in the last demde and are proiected to become minority majority states early in the next century (1.1.5. Bureau of the Census. 1993). In die 1990s. lawsuits brought by universities and school districts in south Texas threatened to force the reapportionment of funds to schools and public institutions of higher education mugeley. 1991-, “School Finance Ruling; Plaintiffs Hail Latest 1i'ictory in School Case." 1990]. In California, state laws prohibit schools from enrolling undocumented im- migrants, and another proposition that was recently ap- prosred by voters effectn'er outlaws bilingual education. Yet as in the 1960s, the discourses of the current literacy crisis are seldom referenced to issues of race, language. and politics. but instead to higher toned concerns about the global economy and the nation's future position in it [e.g.. Christie, 199?; Marl-cley. 199?; “Praise and Criticism for Clinton Education Policy." 199?). its in the 196435 also, it is discourses of scion!ch that is. of research conducted using experimental and quasi-experimental controls—that legitimates the making of policy and the Writing of literacy cuntculums within these states. Newspaper articles appearing in major dailies in bout California (Gelatin. 1995'. Jacobs. 1995} and Texas (Markley. 199?; Woo. 1991') in the 1990s that sup- port a return to phonics- snd skills-based instruction have advanced the same argument as that made by Stanovich (199551994) in a teacher practice journal dur- ing the same period. This argumem is that. whereas the arguments of preceding curricular approaches were based in the romance of unsubstantiated stories about children acquiring literacy naturally, the linear rationality of the Scientific evidence will inevitably lead those who use Their minds and not their emotions to a very different act of conclusions about the reality of litinacy acquisition. While the scientific Evidence these calls for reform invoke originates in the classic studies of the 19605, more recont published research and the curriculum poli- cy documents of California and Texas ground their argu- ments in a phenomenon first described iii a 191% report by Liben'nan, Shankweiler, Fischer, and Carter. and now commonly known as phonemic (or sometimes as phono- ligicrah awareness (PPM. In that research and in a series of subsequent studies conducted in the lil'i'll'is and 19805. statistically significant correlations were found between the ability of young. preliterate chiidmn to segment a word spoken to them by a researcher into component sounds. and their success as readers-“that is. 3.5 fluent and accurate decoders of print—later in school. It has been primarily to this research and its reviews that cur- riculum writers and policy makers in Texas and Califomia in the late 19905 have looked for the justificas tion of their efi'orts. l'iilhen the evidence is laid out in this cold. appar- ently factual manner. there would seem to be only one reasonable way to interpret the meaning of PM for early reading insuuctioo. and so the curricular initiatives of California and Texas would appear to follow complemly and directly from the facts as they are presented. 1r'et ei- ther when the full argument is examined in close detail. or when the resulting curricular changes are viewed widain their full demographic and instructional contexts. the reasoning for these states' curricular shifts becomes more suspect. for Five reasons. First. to fill all the links in their chain of argumentav tion—to explain, for example, why minority and low—so— cioeconomic status (SE51 students tend to demonstrate less phonemicr’phonological awareness than their white or middiesclass counterparts in testing situations—read- ing researchers have been forced to rely upon arguments whose assumptions about race. class. and the home lives of children are open to question on multiple fronts. And second. reliance upon a very narrow range of research, coupled with the exdusion of equally cbiective but con— tradictory evidence. and in California the strong~arm tac- tics of the legislature in banning the use of certain terms associated with whole language in inservice programs {California Assembly Bill 1036, 199?), suggest d'tat policy makers and curriculum writers in these two states are not at all as sure of the Iegitirrtaoyr of their positions or the persuasiveness of the facts as they claim to be. Taking a Haber-master: approach My purpose in writing this article. then. is not to challenge the validity of PPA’s legitimacy as a construct. ‘I‘wo states‘ reading initiatives 26?: —-—.—.—.-.—.——..._._—__n_—__ but rather to investigate and to interrogate its current in— terpretation in scholarly publications and its subsequent use by writers of literacy curriculums in two states. True to die linearity of its own precepts. the path ofthc dis- course that supports the current curricular documents of California and Texas is clearly traceable from reports of experimental studies to reviews of mat evidence to die curricular documents themselves. Both states' documents make a point of declaring themselves to be research. based. both include lengthy bibliographies consisting of scholarly articles and research reports. and the California document routinely includes citations of research articles and related raviews within its text. One analytical approach that is consistent with the theoryrresearch into practice argumentation characteristic of FPA research and its resultittg curricular policies is sugv gested by Jurgen liabem'ias's Theory of ICommunicative Action (198?. 1993: see also Carspeckcn. 1996.- White. 1938}. In that dteory. Habermas (1995) argued that as an overall project. the goal of modernity—to produce a hu- man society based on the principle that each person is an autonomous member of equal worth and possessing equal rights—is not false and misleading. but as yet in- complete. Completion of modernity's goals for Haber-mas depends not on abandoning or muting the possibilities of rationally coordinated social action. but instead on carefully defining what it means to be terrorist. and on delimiting the parameters of rational discursive action with respect to a given propositions sphere of concern. Actions taken on the basis of rational discourse depend on a two-part argument. First. I—Iabermas used the speech act theory of Austin (1925} and Searle (1969} to argue that to speak is also to act. and that the truth claims of any utterance can be determined to be valid or invalid through an analysis based on formal yet pragmatic standards of rationality that all communicatively competent speakers of a lan- guage must necessarily possess. To be conununicativefy competent IWithin a conversational exchange means that a shared predisposition exists among all parties to strive to understand each other. To be rational in one’s ap- proach. then. means to give reasons in support of what one says and does and to urge odters to do the same. in order to open all reasoning processes to public scrutiny. that is. to critical analysis. To Haberrnas, the very ubiqui- ty of giving reasons also demonstrates its communicative and political significance. because in this approach rea— sons are more than mere excuses for actions a speaker has taken or worrld like to take. Rather. they become in- timately tied to the speaker's actions. so that unless someone is deliberately and willfully lying. reasons given can be construed as fonnd‘atfomt to actions taken. Thus for Habermas. who claims membership in the Frankfurt School (see Horkheimer Si: Adomo. 1514431995}. the scientific discourse used by Nazis or Stalinists or Maoists. or by eugenicists and social Darwinists irt Britain and the United States {Gould lacrosse; Hofstadter. 1944-, Willinsky. 1993). did not merely excuse atrocities, it was an effective and necessary part of their production. Habermas's contribution to this critique of modernist rationalism has been to argue that had there been at the time a public forum in which such obiective scientific evidence and its relation to social policy could have been scrutinized. the poiicies themselves might have been shown to be illegitimate. His argument does not suggest that atrocities would not have occurred, but rather that had their legitimat'rng discourses been pub- licly challenged. their political momentum might have stalled and that this would. in turn. have lessened and shortened the policies' efficacy and their atrocities‘ dura— tion. The second part of Haberrnas's theory turns on his proposition that how reasons are determined to be valid varies according to whether statements are made and reasons are given about phenomena that are objectt't-efp-ftysfcnf tie, that have multiple points of ac- cess and can be cross-validated as facts); that are nonna- trie-es'ainotr're (Le. that derive from human consensus developed through rational conversational exchange and that speak to what sitcom“ or page: to be).- or that are sitar-active lie, that the speaker has privileged access to and that must be validated by the logic of their otvn ctiv teria). Each of these three realms requires different crite- ria to determine the validity of a statement. Statements made about objectiverphysical phenomena must be ref— erenced to evidence whose existence is verifiable by multiple observers using procedures that render social values and variations in the perception of individuals ir- relevant. The validity of normative-evaluative statements is testable by whether they conflict or are in agreement with set principles that govern relations among. and the rights of. individuals within a society, a process of vali- dation that presupposes an open forum for such discus- sion that is free from coercion. Finally, the validity of subjective statements is testable by the extent that. when questioned. a speaker can demonstrate the sincerity of her or his views. Subjective statements may be influ- enced by normative-evaluative views and may use objecs tivcly referenced facts, but their subjective validity does not depend on either: thus. one may repudiate the social values that a work of art instantiates or challenge the li— cense it takes with objective reality. yet Still Understand it as a valid artistic and personal statement. According to Habermas, it is the capacity of corn- p-etent speakers to categorize and apply one of these 264 MEG RESEARCH QUARTERLY Julyi’AngnE-tfscptesnhu 1999‘ 34!?! three different and separate tests of validity to their and others' propositional Statements that makes the rational coordination of actions within a society possible, and makes the project of the Enlightenment still achievable. For example. in the case of reading research. it is tlte co ordination of objective-“physical and normative tests of validity, rather than the strategic subordination of nonna- tive to obiective concerns {or vice versal. that makes the goal of communicative action and social progress possi- ble. This is a critical point. because as White (1963) not- ed, it provides a powerful critique of, and reply to, the cognitivevinstrumentalist project of positiviSm, and panic— ularly to "the claim that the validity of science is inde- pendent of any normative commitment on the pan of the scientist." a claim that “also sanctions an understand— ing of politics in which rationality refers only to the ef- ciency of the means for carrying out individual and collective ends, not the ends themselves" tip. ZED. In other words. before objectiver validated facts can be mustered to justify an act within a rational, just society, they must be coordinated within a sphere of ac— tivity constituted by a set of presuppositions not about establishing existence but about establishing tightness. in this case. the fundamental test of validity is not is spent- bie? but Should t'i bedded? and, in support of the latter question. What are all the (and my} reasonsfur acting? and WEI-at are at! the {foreseeable} consequences? This is another critical point. because. from a [-Iabermasian per- spective, to be objective in one‘s. approach to examining human social behavior does not imply that one should assume human social behavior is governed by the same sort of laws that apply to the physical world, or that one should take all reasonable precaution to absent oneself from the design and then declare objectivity achieved. Rather. it means to take full account ofonc's motives and normative assumptions about human behavior at all phases of the research process, not in order to eliminate all possible validity threats. but rather to take inevitable. unavoidable assumptions into account in the analysis of data and in the conclusions about others that are drawn from it. In Haberrnas's critique of modernity, then, it is not an overreliance on the possibilities of rationality or hu- man communication that is the problem. Rather, it is {a} the failure of all parties involved to seek understanding of one another‘s positions,- and (b) the misappllcation of the criteria and practices of one realm of being (Campeclten. 1996] to anodier realm—most notably pos- itivism's misapplication of techniques used to discover and validate knowledge about physical reality to the on dieting of the social, or nonTtativerevaluative, realm—that has prevented the Enlighteth goaLs of a just and ra— tional society from being realized. Such actions, Habcrmas argued. are not communicatite because they are not oriented toward a clear understanding of the is- sues involved on the part ofall communicating parties. instead. they are strategic. that is. oriented toward the achievement of a goal. Eitl'lET through conscious manipu- lation or through unconscious, "systematically distorted communication" (Habermas, 193?. pp. 3-52—3330 about the issue under discussion. his IEatspeeken (1996) noted, moreover, when strategic speech acts substitute for com— municative ones. they frequently if not always accompa- ny and justify the coercive use of power, sometimes through the use of physical intimidation, but also less overtly through control of access to resources and influ- ence. Habem'ias maintained that the reason strategic ac- tions may gain the upper hand in the making of policy is due not to the failure of rationalism, but rather to the failure of a society to establish and maintain institutions. such as a free press or other sites where the coercive use of power to support one's statements is disallowed. and so where the merits of a con'tmunicant's statements can be categorized properly and then fully. publicly, and ra— tionally tested. Because Haben'nas's Theory of lCommunicative Action does not dismiss the existence of an objective realm that can be consensually named, and also because it argues that the improvement of human society through rational processes is both achievable and necessary. a Haberrnasian approach to analyzing curricular policies that claim to be based on objectively validated research evidence would seem to be fundamentally sympathetic to the project and perspectives of these policies“ propo- nents. Yet it would also seem to provide a powerful in— strument for investigating the validity of curricular approaches to literacy instruction that claim to be based on objectively referenced facts, for it acknowledges that the criteria for determining what makes good science and what makes good irisn-ttctt'cncipmctics and gone' social policy are not the same. What is held to be objectively, scientifically valid can only be applied in nonnative, so- cial spheres of action after it has been tested by that spherc's very different criteria and by means of fully public. and open. discourse. Procedure To make such an analysis, I have conducted 3 rev view of the publications that claim to base their propos- als for action on the findings of phonemicfphonologieal awareness research. An investigation of these claims could take many different loci {c.g. an analysis of the methodology of PPA research, or an analysis ofthe me- dia coverage and politics of reading policy. a la Taylor. 1953B). But in this article 1 focus on the claims. of the Texas and California documents that fa]I their policies No states‘ reading initiatives :65 are based on the findings of objectively validated re- search, and that (it) these policies are motivated and cor ordinated by concern for the reading achievement of children in those states. To this end, the task of this analysis is not to sur- vc'y the entire corpus of the PPA and related reading re- search and pronounce it objectiver valid or invalid. but rather to examine the validity of the reasoning processes that lead from claims of obiectivity to curriculum poli- cies. This project has the single goal of determining whether. and to what met“. the rosearch-uito-praetice claims of the two states' policy documents an: validated by the evidence they cite in their support. To achieve this goal. I worked backwards from the Texas and California curricular documents. noting which publica- tions were cited in both bibliographies and which publi- cations were cited most frequently and most extensively in. the California text (there were no citations in the Texas text. although a bibliography was included). Two publications in particular stood out: a book-length re- view of PHI. research and its implications for phonics! skills-based instruction. Beginning to Rend— Thinking and teaming About Print (Adams, 1990). and a maior, fre— quently cited literature review of PPA research and its theoretical implications. "Matthew Effects in Reading: Some Consequences of Individual Differences in the Acquisition of Literacy" {Stanovlch 1935). An analysis of the Frequency and extensiveness of citations in these twb literature reviews, when checked against the bibliograv phies of the cumcular documents. yielded a list of 10 maior, influential reports of original PPA research (a par- tial exception is Ehri. 193?. which reviews several of Ehri's own previously published studies). The signifi- cance of these research publications was also verified by their citation in publications outside those selected. as well as by the citation of the earlier publications in the later publications that were selected. In keeping with Habermas's “theory of ontological categories as universal features of human communica- tion' (Carspeclten. 1996. p. (11}. these 14 documean were initially categorized by the type of evidence used in support of their claims. that is, by claims referenced to the objective realm of experience. to the snhjecrt'te realm ofexperience. or to the Pronounce-eminence realm (or, as Carspecken. 1996. put it. to "the world“ [p. 65]. "my world“ lp. Til]. or “ourworld” [p. 33]}. With few excep— tions, the 10 documents reporting experimental and qua- si-esperirnental research on PPA referenced their truth claims to objective criteria. The first consideration iti evaluating these claims was the extent that, in fact, the interpretation of PM remained wholly obiective—that is. depended solely on empirical evidence and nothing but empirical evidence—or could be determined to be influ- enced by normative assumptions about the nature of language and differences among children tested for PPJL To the extent that findings and interpretations were shown to be normativer influenced, their validity was evaluated by consideration of their tacit claims to social tightness. that is. principally by consideration of their orientation toward egalitarian purposes. Placing the claims of the two reviews of PM. re- search in a single category of criteria proved more diff- ctilt, for while they closely relied on [actuality in forwarding their claimsI their main task was to interpret the meaning of research findings about PPA and early readers' levels of performance in such a way that they have application to instructional practice. These interpre— tive acts are alternately supported by claims referenced to objectivity. to their audience's sense of what is nonna- tiver self-evident. and at times to their own experiences and feelings as parents and concemed citizens. Discursively. they were the most complex of the three types of documents to be investigated, yet they were also the publications most frequeme and extensively cit— Ed in the Texas and California curricular statements. The curricular documents from Texas and California, on the other hand. while referencing the claints of research re- ports and their reviews. primarily argue not about what is. but about what strontd be. For this reason. their claims were examined in light of nonnative-evaluative criteria of what is right and first—a task that also involved conv sideration of the extent to which these documents‘ inter— pretaticms of the findings of PPA point, as the documents claim. to only one conclusion about how to teach begin- ning reading. To analyze the 14 doCuments. I used a process of textual analysis adapted from the more comprehensive, multistage methodology ofethnographic data collection and interprecation proposed by Carspeclten (1996}. which integrates Habermasian principles of rational dis.- cursive action with principles of hermeneutic inferencing as discussed by Palmer (1969. cited in Carspeclten. 19915). First. the r'iioc‘a strainer}r utterances of each text (that is. propositional statements that materialize and are meant to actively advance a speaker's or an author's claims} were isolated in order to construct the primary record of the analysis. In actual practice, this meant read- ing through each text multiple times and underlining claims that were original to that tract, while noting the evidence cited in support of those claims. Concurrent with this process, the ontological category (obiective. nunnative-evaluative. or subiectivelI that each truth claim invoked was identified for later verification and analysis. These propositional statements. when represented in tabular and summary form. constituted the primary record (Carspeclten, 1996) of the claims presented by the 266 Resume assessors QUARTERLY Julyrauguavseptember 1999 sits lt‘i research reports, 2 reviews, and 2 Curriculum docu— ments. They are visible in this study in Table l and in the summaries that precede the artalyses of the reports, each review, and each con-iculum document. a partial exception within this stage was the treatment of Adams‘s (1990) review, as its tililili-page length threatened to over— wheltn the signficancc of the other 15 texts. instead. for its primary record I composed a summary of the text based on accumulated readings and annotations of the entire text as well as on class notes I compiled over 3 sev theaters of teaching an. undergraduate course in which the Adams test was required reading and the focus ofB weeks of class disorssion each semester. in the second Stage of the process, I examined ead‘t claim in the primary record of the combined re— search reports and each review and document to consid- er whether me criteria specific to the ontological category invoked by that claim 1were met. in Campeclten’s (1996] ethnographic methodology, this recomtmctt'te dentists is wide ranging and involves articulating and sorting through all the possible meanings that a gesture, a tone of voice. a comment, or a more exrended set of actions and interactions might have for the participants in an exchange. Conducting a reconstructive analysis oi" the 14 texts was in many ways simpler than the process Carspeciren described, however. for three reasons. First, the range of data that I analyzed was considerably nar- rower than that collected ethnographically. Second, the arguments contained in the texts were more deliberately constructed and oontexmalized by the authors than are naturally Occurring speech acts. And third, the focus of my investigation was much narrower than most ethno- graphic analyses. ln the analysis I conducted, I returned to each illocutionary point noted in the construction of the primary record and looked to see if obiective claims were substantiated by purely objective evidence, or whether and to what extent nonnative-evaluative as- sumptions figured in the author's reasoning process. If nonnative-evaluative assumptions did figure in his or her reasoning, [ looked to see whether their presence was acknowledged and whether these claitns could be vali- dated by normative-evaluative criteria—omens based on assumptions of human equality and the promotion of so- cial justice. .-'is proceduralized as this analytical technique is, however, it is still a henneneutic process of inference making (Palmer, 1969), and its efficacy depends, in the end. on the researchers own communicative compe- tence. which is developed largely through the rev seart‘her's lived social experiences. in Carspeciten's {1996) methodology, the process moves from relatively low levels of inference making (that is, inferences about the immediate situation) to higher. more distanced levels of inference that become the researcher‘s point of refer— ence, in macrotheorett'cal terms, when speaking to other researchers and social theoreticians_ Carspeclcen also specifies a third stage of critical ethnographic analysis, in which the researcher conducts interviews, or member checks, with the participants in his research, both to check his own interpretations and inferences. and to “democratize” (p. 155) the research process. For masons ofpractioality, because written, published texts are more shared. complete rilaiects of analysis than are ethnographers' field notes of ephemeral social interactions, and because there is a tacit claim within each of these texts that their meaning is transpar- ent, I did not take the step of contacting or interviewing each author. Instead, drafts of this study or of its pro- posed analysis and tentative findings have bceri submit— ted for blind review twice (once to Division B of the Anterican Educational Research Association and once to the editors of RRQ). The process of connecting infers-noes about one's data to larger social issues and then to social theory itself constitutes stages four and five in Carspeciten's critical approach. These stages are visible in the concluding paragraphs of the analyses of the reports. reviews, and curriculum documents, and in the study‘s conclusion. Resufls Dhiocflve chime: Readings of 10 research reports Table 1 lists the maior findings about PM. that ap- peared in the 113 selected research reports on PPA. Secondary claims mentioned by the authors, that is, dain'ts attributed to other research reports, have been ex- cluded. as have any claims that did not directly pertain to PFA and any statements made by the authors that were acknowledged by them to be speculative in natune. All of the findings listed were supported by objective measures, that is, through subject counts, experimental designs with Fairly rigorous validity checks, and the use of statistical analyses to test for significant differences among groups. ‘v'ery briefly, one of the most striking observations about these studies is how robust their principal findings are and yet how complex and open to interpretation their instructional implications remain. All it] studies are in agreement about the definition of PPA as a construct, 9 studies concluded that there is a causal connection be— tween FPA and the ability to read fluently, and 3 studies found that PM. is in fact a prerequisite of fluent decod— ing. Yet ‘i of the studies also found that being able to recognize some whole words on sight also contributed to the development of PPA. Ehri (198?), citing her own Twnr states' reading irfltiatives 261' Table 1 Selected oils-jectinrel'.r referenced claims of FPA research reports {)h-lectiuety validated claims {in paraphrased language} a 5. Claim Bradley it Bryant. 19H}. 12!. 419: CunninghaJ'l'L. 1990. p. 429: Ehri. 1937'. P— 215: .lI-IEI. 1933. p. 433'. fine] et al.. 1936. p. 245: Libennan et s]. 191%. pp. Em—El'fl. Perfetri et al.. 1%?. p. 284. Tttmne-r tir Nesdale. 1955. p. 1'12. ‘t'elltttlno .i Scanlon. 198?. p. 5311; Yopp. 1933. p. 115 Juel. 19158. p. 45?: Ferrari etal.. 1513?. pp. jun-.5111; Velluunosi Scanlon. 1931'. p. 329; 1t't'inpp. 1938. p. IN Yopp. 1983. p. IT4 Cottth definition of PP}. a. PM. is the ability lo manipulate 1he component phonemes of oral language in some way (eg. by segmentation. deletion. or Mending) Is. As a construct. PM is characterized by multiple Factors c. As a construct. FPA consists of two factors. Compound {involving tasks tutr deletion) and Simple (involving tasks of segmmttation. isolation. and blending) Relauionsl'lip to reading achievement a. The abilin to segment wards into phonemes is the measure of PFA m highly mlared with reading achievement (that Is. mat-d Juel. 1933. p. 440-, Perfeui et at. 195?. pp. SIT—313; Veilnt'tno s SCanlm. 198?. pp. .525. 342. 3151: Yopp. 1983. p 1'14 identification} h. It causal relationship between a prereader's FFA and later reading Bradle't.r at Bryant. 1956-. p. 419; Cunningham. 1990. pp. 451—455; abllrry is reported Juel. 193-3. pp. sclfl. I114: Juel et al.. I935. pp. 215-255: Perfetti et al.. 1911?, pp SIG—51?; T‘Lwner t1- NesdaIE. 1935. p 425; Vellutino at Sanlon. 1961:2111 359. 55?: YGPP. 1965. pp. tint—trs Clothingham. 1990. p. 433tjuel et al.. 1933. p. 245-. Feri'etti et al.. 1%. pp. 505. 3051: Vent-lira: s: Scatflnn. 1911?. pp. 53-11. 5'55. 359 Cum-ringham. 1991]. pp. 436-437: Ehn'. 193?. p. 26;]uel. 1968. pp. HS—flfiijucl Hal. 1986. p. 245; Perfem 21311.. 193?. pp 3-15. 31?; Tunrrier & Nesdale. 1935. p. 425-. 1tl'ellLrtino & Scarllon. 1981'. p. 5'51; Yepp. 1933. pp. 114-115 End. 193?. p. 2E! c. The relationship bemeen PPJL and whole-word knowledge is no some extent reciprocal cl. Fluent decoding depends on some prerequisine hinndedEe of PPA e. PPA “may be part oF learning to read rather than something thasz tleuelopeel beforehand" Instructional implications a. PPr't can he taught Bradle].r & Bryant. 1935. p. 4211;. Cunningham. 1990. p. 449-. Ehri. 1937‘. pp. 26—F:Juel et aI.. 1935. p. 249-. Liberman er a].. 1931. p. 21!}. Petr-ens H 11.. 1937. p. 313; Tunrner & Ntmdale. 1955. p. 425: Vellulino s Sc'anIDtI'L. 193.1. [1. 35] Bradley & Bryant. 1953. P— 419-. Cunningham. 1990. pp. 453.. 4-“: ‘r'eltuu'no & Seanien. 193?. p. .160 h. Teaching prereaders to analyse Iwords Lrtro component sounds is mere powerful instructinnally than is teaching prereaders to synthesize phonemes into words 1:. Deconiextualized direct instruction Ln suhskills of PPA dorm not transfer to readingt'tleeotling tasks as well as enmenualized, analy‘tlcal instruction d. Widrotrt specific PFr't analysis tlaining as a component. mediod of reading it'eeutnon does not afieet development of PM 2. Film-tics insttucntm is net effective before children have acquired PPA Clmningham. 1990. pp. MEL-=14] Tunmer at Nesdale. 1935. p. 425 Joe] et al.. 1933.. p. 246; Ubenmn er al.. 1994. pp. 210-21] _ Tuan-r wk Nesdale. 1985. p. 425 React-archers ail-arenas of laslt and acoustic problems a. Performance of some tasks requires ttaining in w.- taslt. apart from Pile Etu‘inr'trgham. 19911. p. 433-. Perfetti et al.. 1911?. pp. ass-sou h. Difficulty of analy5as task depends on. nonrigurry n! phcimsmes Pei-fem :1: al.. 193?. pp. 295—500 1‘. Tests that use digraphs (El-phoneme corrtrirratinns) are unreliable Warner & Nesdale. 1935. p. 425 PHI. and prereaders' ethnicity and SES :1. EI-l'lrlit‘il}.r arrth'or 5E5 oi subjects is reported Joe]. 1933. p. 439. juel er al.. 1955. p. 2-16. Eben-nan at al_. 1911. p_ 211-1: Perietti etal.. 1915?. p. 292: Yopp. 19133. p. 165 Bradley 3: Bryant. 1935-. Cunnlngham. 1990. Tunmcr iii Nesdale. I955: Vellulim & Scanlion. 1957'. p. 525 11.11:]. 1983. pp. “Hie: libel et al.. 19%. p. 219*. Perfetti et ai.. 1981'. P- 292 h. l’ifltnitity and 51:3 or" whit-us is not reported (but urban-“suburban! rural differences meruioried h}- ‘v'ellutlnn and Seminal c. ErJ'micinr is significantly- con-elated with FFA in beginning readers: I:le was an implicit factor in Perfet'u et al. (193?. 12 of 26 reads-rs in the high reading group were white). and inJLsel {1938] research (Hohn dc Ehri. 1935) and that of Bradley and Bryant (193:1) in which speiling knowledge and letter recognition in words made a statistic'all'].t significant dif- Fe:ence in children's ability to segment words into phonemes. argued that “phonemic awareness may.r be part of learning to read rather than something that is de- veloped beforehand' (1931', p. 2151. 111e- instructional applications noted in these smel- ies are even more complex. although all the Studies found that PPA can be taught and E- argued the advan- ass assume RESEARCH QUARTERLY Jflyffiugustffieptemher 1999 W3 tag-es of its explicit instruction. discussion about the form that instruction should take and its relationship to other aspects of reading instmction was less convergent. Generally, however, analytical procedures that were con— cunent with the reading of actual tests or that empha- sized the relationship of phonemic segmentation to the alphabetic principle and the application of that knowl- edge during reading were more helpful in the dEveiop- ment of PPn. and in the transference of PPA to decoding words than were procedqu taught apart from the read. ing of texts or that emphasized the blending of phonemes to form spoken words. Howeverr none of the studies examined claimed that the teaching of phonics substituted or compensated for a lack of PPA. To the contrary. in the discussions of their results, Juel, Griffith. and Gough (198.6} and Liberrnan et al. [19%) itnplied and T‘unmer and Nesdale [1985) direoly stated that phonics instruction is not effective in teaching children to read who are not already phonemicallyrphonological— ly aware. A problem mentioned in several of the studies (Libel-man et al., 1554; Perfetti, Beck, Bell, lit Hughes, 198?; Tunmer d: Nesdale, 1935} is that because PM is a construct that can only be determined indirectly through perfon'nance of specified tasks, a possibility always exists that what is being measured by any single taslt is not the construct. but the subject's understanding of the task and the cognitive strategies he or she uses in its perfor— mance. In the case of PPA, this is largely because even though Cultures with phonetically written languages opv erate on the assumption that language is composed of individual sounds, in actuality those sounds only exist in contiguity with other sounds within a stream of speech-r they cannot be spoken discretely. its an example of this, try making the Is“ sound by itself; then say the word sum and note the difference in the position of your lips in the two instances. Or try making me Edi sound in isolaan and note the aspiration that accompanies it {dub}: but then say dog. Did you hear any aspiration when Ids" was pronounced within a word? Ctr try saying DfdJ-Du eat? as you would in normal conversation and pay attention to the sound the second r'df in did makes in combina— tion with the rw in you. in the dialect of English I speak. the Ida“ and the r'yv’ combine to make the all sound llit-'liicl1 in languages like French and Arabic is heard and spoken as two sounds. fdrand the .in in azure); but perhaps you speak English in a different way. Neverthe- less, we probably agree that did has three phonemes and you has—one? or two? These are the issues to which Libemtan et a]. fight) alluded when they noted. "IT'lhe reading teacher who LIng the child to blend b—a—t (erasures) into the appropriate ntonosyllabic word bar can only succeed in teaching a child to say bunnies. since "phonic segments are encoded at the mode level into larger units of approximately wllahic size... Thore is, then. no acoustic criterion by which one can segment the sound {of a word} into its constituent phonemes” (pp. 203-204; see also Perfetti et al, 193?; Tunrner iii: Nesdale. 1935}. The problems encountered in attempting to inventory the constituent sounds of any given lan— guage (that is. to idectify a Table of Elements) and de- scribe their behavior with any precision (that is, to produce a phonological chemistry} have, in fact, proven so intractable that since the late 191505 phonologists have waged a great debate of their own about the theoretical validity of the phoneme as a construct (Chomsky St I-Ialle, 1968). One side dismisses the traditional idea of the contrast-based phoneme as ‘hiemonian" (tense-n. 1993, p. 3} in its assumptions, and the other defends it as *not contradictory but complementary” (Lass, 19%, p. 212} and still fundamental (Geigerich. [992) to the ab- stractions of current phonological theory. Regardless of which position individual phonolo- gists take, there is consistent agreement in the field that each language. and indeed each dialectical variation of a language, is a unique system of sounds, albeit a system of phonemes and rules of combination that overlap with other systems. Multiple studies also indicate that people's phonological systems are set before puberty (and per— haps as early as 6 years of age) and remain relatively fixed throughout their lives (Flege at Fletcher. 1992-, Long. 1990‘. Ciyama. 19m; Patltowski, 1994-, Scovel, 1933‘. Thompson, 1991). Bilingual speakers who acquire a sec— ond language during early childhood eventually develop relatively but not completely separate sound systems for the two languages. whfle those who acquire a second language in later childhood develop a compromised sec- ond system and nearly always speak with a perceptible accent, regardless of their le'vel of proficiency (Flege lil- Port, 19m; Yayas, 1998}. a compromised second phono- logical sysIEm results in first-language interference not only in the production of sounds but also in habits of auditory perception that are part of one's primary phono- logical system—that to a speaker’s ability to distin- guish the sounds of his or her second language in the same way as a native speaker would. Though perhaps not as dramatic. the phonological systems of a particular dialect are also set early (Bough, 1983i; Melmed, 1W1; van Keulen, Weddington. lit IJeBose, 1996}. Differences in the phonological nonns to which members of a particular speech community he— come accustomed are such a common source of confu— sion for speech pathologists who work with bilingual and dialectically different children that textbooks in that field repeatedly warn against misdiagncsing differences in a speaker's phonological system as the indication of a Twostaiaes' readinglnlflativies 2.69 speech disorder (see tiiecharn 8; Willbrand, 19W“, van Keulen Et al., 1998; Yavas. 1998). The implications of these warnings for researchers who would try to mea- sure the PPA of children who do not speak the same tli— alect or first language as themselves are clear. Yet none of the it) research reports here, nor any research on phonemic awareness that I am aware of. makes any mention of these issues or of precautions tat-tee in the collection of their data when working with linguistically different populations. A second task-related problem with determining a child‘s PPA is that the phonemes and syntactic fortris speakers use in expressing themselves not only carry cognitive meaning, but also situate a speaker in a panic— ular social, cultural. and political space vis-a-vis the per- son to whom they are speaking, often without the speaker's awareness and despite his or her best attempts to reach understanding itlvitl'i an audience. Numerous so- ciolinguistic studies conducted in the 19?0s and 1930s that used methods of triangulation and analysis that were as objectifying as the procedures used in the PPA studies cited pmiously established the extent to which seem— ingly innocent and transparent spoken forms used by teachers—for example, simple questions directed to an individual student about a story or direct commands to do something—were interpreted by their students as threatenineg direct or otherwise inappropriate thoggs, 1935: McDermotr. 19%; Philips. 1955; Hist, 19ml. Perhaps the most classic instance was described by William iabov {19?2}. a phonologiSt and sociolingulst who studied the speech patterns of African American adolescent and preadolescent males in the inner cities of Philadelphia and New York. Early in his research Labov realised that the responses young African American boys gave to interview questions posed by smiling, friendly white graduate students were qualitatively and quantita- tively different from those given when the boys were in- terviewed by an African American male who was known and respected by them. When they were interviewed by whites, the boys' responses were short, evasive, and ap- peared ungrammatlcat and nonsensical; when inter— viewed by African Americans the boys responded in a series of extended exchanges that demonstrated their in— tellectual command, not only of the grammatical and ideational content of the conversation, but of the com- plex conversational norms that govemed their exchanges (Labor. lQfl). what is most relevant about Lahov‘s analysis with respect to PPA research. however, is his conclusion that the responses of African American children to white interviewers were not caused by conununicative incom- petence. but were “rather the result of regular socio- linguistic factors operating upon adult and Child in this asymmetrical situation“ to. 20?]. This is a situation where the child believes that ‘anything he says can literally be held against him...(and where he} has learned a number of devices to avoid saying anything in this situation" lip. 206}. Lahov went on to report that in many instances white researchers were often blind to the effect their eth— nicity had on the interview situation and concluded that their subjects were really as inarticulate and as "linguists cally and cognitively deficient" (p. 2%} as their data sug— gested. He singled out the work of Bereiter and Engelrnann {1966}. who characterized the language of the African American 4-year-oids they interviewed as "a basically nonlogical mode of expressive behavior" [cited in Labov, 1912. p. 205] as a prime Example of this phenomenon. Inbov's findings and discussion have clear implica— tions for (white, Huddle-class} PPA researchers who at- tempt to get minority and low-SE5 children with whom they have only a passing acquaintance to perform com- pleat phonological tasks like tapping out the number of phonemes in either real or nonsense words spoken to them out of context. Yet only 5 of the IG studies ana- lysed in this article report the ethnicity or social class of their subjects, and none report the researchers’ ethnicity or socioculle background. None of the studies menv tions issues of dialectical differences between them or their subjects, even though diese must have existed, and none reports asking their subjecrs to explain their re— sponses or making any attempt. to dehrief adults in their community about the responses children gave. Despite the work of iabov and many other sociolinguists, and perhaps more tellingly, in contradiction to PPA re- scarchers' own acknowledgement that their research measures children's awareness of a social convention (phonemic segernentation) rather than an objective, physical phenomenon, PPA reports persist in characteriz— ing language in general and PPA in particular as fixed objects rather than as a more consensual. culturally shill— ing construct. Because sociocultural difference is not even men— tioned in half the studies discussed here, readers might conclude that, as in the case of the FirSt-Grade Snidies, it has larger been ignored in discussions of the implica- tions of FPA research for reading instruction. But as the discussions of Adams {19%} and Stanovich (1936, 19914994) in the next section indicate. iust the opposite is the case. in fact, one of the principal themes of read— ing researchers who rely on PPA findings in their discus- sions is that low-SE5 and ethnic minority children demonstrate teas PPA than white and middle-class chil— dren. in conclusion to the analysis of this section, then, i examine the evidence supporting the claims of l of the 2 (out of 10) research reports in which sociocultural back- are assume assesses ouaa'r'satv rutyrAugusvSeptembcr 1999 sets ground was claimed to have an effect on children’s levels of PM. and consoqucntly, as the researchers explained. on their reading achievement. In that study. reported byjuel et a]. (1936; see also the related report of Juel. “988). 129 first-grade children {so ‘nnglo Americans.“ 45 "Hispanic Americans." and 30 "black Americans"; p. 2445) were followed over a 2-year period. During this time data were collected on the stit— dents’ IQs, oral language-“listening comprehension abili— ties, PPf'i. writing, exposure to print (as measured by where their reading group was in the school’s basal se- ries}. and several measures of reading achievement and spelling. In their model of analysis. a regression equation with PPA as the dependent variable and ethnicityI IQ. and stores on a test of oral language as the independent variables found statistically significant negative coeff- ciertts (calculated as "the average deviation from each ethnic group explicitly included in the model from the Anglo group," p. 248} for African American children (".45 die first year and —.29 the seeond, p s .01) and Hispanic children (—26. .tJ s .05 the first year and —.1S|. .o :- flfi the second}. From this model they concluded that three factors—ethnicity. IQ. and oral language—affected a child's level of PPA. which, in conjunction with a child's level of print exposure, in turn determined the autumaticity with which he or she learned to recognize words, to spell. and eventually to comprehend. To explain the negative relationship benveen the ethnicity of the African American and Hispanic children in their snidy and measures of PM. {and so, ultirrtately. to reading achievement}. juel et al. (1986} speculated: Environmental influences, such as rich exposure to the. English language in the home, and their interplay with a child's natural abilities, could cause such dilTerenoes. We suspect that tlte relatively poor reading achieventent of minority children is partly attributahle to poorer phone- inic awareufls of school English due to dialect and sec- ond-language differences, and perhaps to cultural differences. 'Jhc latter may involve such things as "play— ing‘ with words at home through rhyming games, expo- sure to printed Word rhymes as in nurseryI rhymes and Dr. Seuss books (p. 245} interestingly. Jtrel et al. {1936} did allude to the ef— fect of dialect and second—language differences. Yet they made no mention of any attempt on their part to meav sure the students‘ PPA in their own dialect or home lan— guage and so obtain a measure of phonemic awareness that they could claim was free of cultural influence. Instead. they suggested that the digrjlersrrce between the students' own speech and the dialect of English legiti— mated by the school. rad'ter than any objective measure of the students’ PPfi-i. accounted for their low scores. in making this suggestion. they also virtually concede that their results may not have been due to low PPA but to language differences between themselves and their sub- jects. Moreover, they presented no objective evidence in support of their conclusion that the children they studied did not come from homes where “rich exposure to the English language" (p. 245} was present. that is. where children did not play wldi words or have exposure to printed word rhymes. The only evidence they offer about the language and literacy practices of the homes from which their subiects came was taken from infonnal conversations with the children at the end of the study, when "very few of the children said they read outside of school. and even fewer (that is, less than 5%) could tell about stories they had read” (p. 24?). It is clear from Juel et al.'s (1956] discussion of en- vironmental influences above. as well as their descrip- tion of what constituted exposure to print. that the only form of literacy that was salient to them—and therefore that would lilter be made salient in their interviews with children—was storybook literacy; that is, the sort of liter— acy events that middioclass parents stage for their chil- dren at bedtime. Just as clear is that me only sorts of language games they regarded as rich were also those frequently staged by middle-class parents. that the reading of {English} nursery rhymes and Dr. items books. In contrast. multiple ethnographic studies of the actual language and literacy practices of minority and loavx‘SES Families and environments {e.g.. Eoggs, 1985; Heath. 1933: Labov, 19H; Moll, 1992) provide thick de— scriptions of remarkably full oral and literate traditions and practices. albeit traditions and practices that are qualitatively different from the storybook tradition of the middle class. From a. Habermasian perspective. the evidence that juel et al. (1986) cited in support of their explication of the effect of ethnicity on PM. that they reported is not based on objective criteria at allI but on Hormones as» sumptions about What languages and dialects should dominate the school classroom (that is. the ones they Speak) and what they hold to be die right way to sup- port young chiidren's acquisition of school literacy. These are normative assumptions grounded not in obiec- tively validated facts but in untested beliefs about how omers raise their children. In summary. an analysis of to research reports fre- quently and extensively cited in reviews of Iii’fii research shows strong objective evidence in support of Prey va- lidity as a construct. of researchers' claims of a causal link between the ability of prereaders to segment spoken words into component phonemes and their later success as readers, and for claims that PPrl. can be taught. The implications of these findings for instructional practice, Twn states’ reading hiltlaflves 211 however, are less corudusive, although teachil'tg children to analyze words and reflect on their analysis seems more promising than teaching children to synthesize and blend sounds. 1t‘I‘fI'iere objectivity seems to be lacking in these studies is in the failure of all but two to even ac— knowledge the impact of tlte researchers' and their sul‘v iects‘ sociocultural and linguistic backgrounds on their research processes. In the case of one of these. re— searchers failed to distinguish between their own nonna- tive assumptions and objective evidence in speculating about a Fine of causation that links children's ethnicity and presumed home language and literacy practices to measured levels of EPA, and subsequeme to their read— ing achievement. The meanings of PPA research for policy and practice In this section I analyze two publications that, rather than report new findings, review the authors' own and others' research in order to advance arguments about the practical and policy implications of that re- search. I taire the position and I believe my analysis will show that although these texts read as though their con- clusions are the direct and obvious product of research findings. many of the policies and practices that they ad vocate are, in fact, constructed from nonnative assump- tions about human social behavior that. in turn, are supported by evidence that cannot be obieotively sub- stantiated. Matthew Efl‘ects in reading. although "Matthew Effects in Reading: Some Consequences of Individual Differentles in the Acquisition of Literacy“ (Stanovich, 1936) predates curricular reform movements in Texas and California by nearly a decade, many of the claims made in this literature review are discemible in those documents as well as in later work such as Beginning to Read (Adan-ts. 1990). Its overriding theme was that the maiority of disabled readers' problems, and so by exten- sion the majority of their failures (academic and, it is suggested. otherwise). were the consequence of their failure to acquire PPA in their preschool years. Matthew Effects began w'td-t a discussion of the problems associat— ed With establishing causal relations in the field of read— ing disability research. Based on the work of Bradley and Bryant (1983) and others, it proposed a model in which early Failure at decoding and word recognition de- lays the development of automaticiry in word recogn- tion. As a result, reading became a frustrating and unrewarding experience; the reader read progressively less.- Iack of practice led to weak vocabulary acquisition and a failure to acquire "much general information and l-rnoudedge about more complex syntactic structures" (Stanovicl'i. 19845. p. 564}; readers began to fail at other school subioCts that require them to acquire information through reading; and general school failure is the likely result Beginning with the section “Consequences of Reading History and Practice,“ evidence provided by his- torical and qualitative studies is interwoven with experi- mentally based research. For example. the work of anthropologist Jack Goody (The Dnmesttcatfon qfrbe lineage Mfan 19??). of cultural psychologists Sylvia Scribner and Michael Cole If The P39550303}? drifts-meg 1981}, and of jeanne Chall (1985) are cited in suppon of the claim that “the cognitive consequences of the aoqui- sition of literacy rrtay be profound. A few reading theo- rists have warned that we should be giving increasing attention to these types of effects“ to. 3T4). which {Shall argued originated during the early stages of literacy ac~ quisition. These arguments prepare the reader for an expla- nation of the causal relationships among FPa. reading achievement. and poveny. Using a quote from the Gospel of Matthew (borrowed from Walberg d: Tsai, 1933), "For unto every one that hath shall be given. and he shall have abundance: but from he that hath not shall be taiten aWay even that which he hath" (cited in Stanovich, 1986, p. 331}. as its metaphor and namesake, Stanovich argues mat PPr-i is the factor that enables read~ ers who have PPA “in abundance" to acquire even greater cognitive riches. While those who do not, have “taken away even that which {they have)_" This biblical observation is also recapitulatcd in the biological princi— pal of "otgrtrthm-enuironmenr coweiarion: Different types of organisms are selectively exposed to different types of em'irottments" (P. dfill. Thus, children who have rich PPA quickly acquire automaticity in decoding. which produces the desire to read more. which leads them into situations where they are more lilter to be ex- posed to opportunities to interact with texts, which leads to increases not only in their reading ability but also in their oognitive development. and ultimately, to superior academic performance and career success. Conversely. “there are also parasite organism-environment correla- tions that contribute to rich-get-richer and poor-gei-poorv er effects. a passive organism-enviroru'nent correlation is a relationship between the type of organism and envi- mnn‘iental quality that is not due to the organism‘s selec- tion and shaping of the environment" (p. 532}. In other wonls. some unfortunate individuals are born into im- pcrverisited environments where they never have the op- portunity to realize their potential. According to Stanovich. this misfortune is the result of having parems who have inferior genotypes: 2T2 MEG RESEARCH WWII? Julyffiugustfseiltembu 1999 343 The genotypes of a child's parents partially determine both the lusme environment of the child and the child‘s genotype. other passive organism-environment eon-eta- tions are a function of social stnict'tlres: Less healthy organisms grow up in impoverished environments. Blo- Iogicslly unluclcy individuals are provided will] inferior social and educational environments, and the winners of the biological lotler are provided bettIEr en'i-‘I'ronmeols. in. ass} In this way. a narrative is advanced in which children with low reading achievement and low ievels of PP'A (that is. low—SE5 and minority children}. bom to parents who are likely to also have low PFA and poor literacy skills. are disadvantaged before they even enter school. Moreover, once in a neighborhood school, all dieir fel- lows are also likely to come from impoverished bach— grounds. so that “quite apart from the individual benefits of above-average intellectual ability, a child of any level of ability is likely to trial-re bener progress if taught in a school with a relatively high concentration of pupils with good cognitive performance” tRutter, 1933-, p. 19. cited in Stanovich. 1936, p. 533}. From this point the paper of- fers a reconoeptualization of reading disability and sug- gests that linguistically impoverished environments are responsible for poor PPA, and consequently for disabili— ties in reading. which leads. in turn, to global cognitive effects and typically lower IQ scores among the reading disabled. its was the case in the analysis of the it) research reports in the previous section ofthis paper, the argu- ments made in Matthew Ellects are carefully referenced to claims about the objective nature of the phenomena they describe. Yet as was also the case. fust as the text moves from discussions of discrete, isolatable cognitive phenomena to more global explanations linking those phenomena to social consequences, its reasons seem less grounded in systematically collecred and analyzed. obiectively referenced data than in interpretive leapsI and implicit appeals to readers’ beliefs about the tightness of the literacy myth (Graff, 1991). For example, in support. of the claim that literacy has global cognitive effects for individuals, and by extension, for societies, the works of anthropologist tacit Goody (19??) and historian David. Olson (19'??-, IElllson, Torrance, dz l-lildyai'd, 1985} are cit- ed. These scholars do suppon this claim, but their argu- ments are based largely on the sort of conjecture and narrative approach to reasoning that, ironically, author Stanovich {19951994} dismisses as romance in a later publication. The solid empirical evidence that is cited in Matthew Effects. however (Scribnerdr Cole, 1931), dis- putes this claim. in their work with the 1irai, a West African group who learn to read and write their own script outside of school, Scribner and lCole found only specific cognitive differences betvveen nonschooled liter— ate and iiliterate list and concluded that literacy per so did not produce the sort of global cognitive consev quences that Goody and ltillson argued were the legacy of the invention of a phonetic script (although schooling did seem to be a significant factor in the "i'ai's ability to understand and perform cognitive tasks in the same ways as Westerners). Moreover, the claims for global his- torical, economic, and political effects that Goody's and Olson's reasoning—and by extension the arguments of Matthew Effects—trace to the phonetic alphabet have been discredited by the empirically based historical re- search of Harvey Staff {193?}, in which literacy levels were found to follow or at best parallel economic, politi- cal. and historical developments throughout the history of the West. Although these points may seem to be no more than slight oversights, from a Habennasian perspective they do call into senous question the paper's global, ob- jectively referenCed arguments about the social and so- cioeconomic effects of PPA. A more serious issue may be the way the process of social reproduction as a con- sequence of PPA is explained. The use of the term geno- ppeis particularly problematic, for while that term does not necessarily correlate levels of PH! oath racial back- ground, il does clearly and strongly suggest that PFA is heritable or at least that it follows from ord-ier heritable traits. and so is to some extent blogenetic in its [Whale tion. Yet nearly every author of every published PPA study makes it clear that PEA is not a natural part of lan— guage but must be acquired through a social process. To explain this apparent contradiction. studies are cited in- dicating that regardless of his or her ability level, a child will perform berter when placed in a school and class With above-average poets d'tan when placed in an envi- ronment composed of peers of his or her own ability level or below l[Fir-bar, 1936; Jencks. 1W2; Rutter. 1935). ‘lhe quality of an individual child's environment, accord— ing to Matthew Effects, would appear to be the more critical variable influencing that child's success, in all that term’s social meanings. But the quality of that environ- ment—that is, the quality of the individuals who com— pose it—is generically influenced. Thus. a chiid of any ability born into such an environment is an advantaged organism because of the superior envi— ronment and genotype provided by the child's parents. The parents, similarly environmentally and genetically ad- vantaged, are more lilter to reside in a oornn'tunity which provides the "concentrath of pupils" that. via the incle- pendent el'fecrs of school composition. will bootstrap the child to further educational advantages. (p. 333} 'I‘vvostates’madingirdflartives 273 although little mention of social class is made in the text of Matthew Effects. the implication remains that the dis- tinction bETWEET'l communities with the right concentra- tion of pupils and those without is mainly socioeconomic and that there is a relationship between SES and intelligence, as indicated by school perfor- mance. This point is underscored in a summary quote from Rutter {1983}: “Nevertheless, the implication is that there are considerable disadvantages in an educational system that allows suclt an uneven distribution of chil— dren that some schools have intakes with a heavy pre- ponderance of the intellectually less able" (p. 20, cited in Stanovlch, 1986, p. 584}. According to Matthew Effects, the selection processes that produce social Stratification are not histor- ical, political. or economic. but instead are primarily the result of reciprocal relations berween genetically advan— taged or disadvantaged forebears and the environment they have produced as their legacy. The snidy does not argue that such reproductive processes are inevitable— since they might be addressed through "sound social policy in education" (p. 58-4}, or more individually, since a chance always exisrs for the odd poor child who is bright to he placed serendipitously in a school or class- roOm filled with a preponderance of high—ability. high— achieving peers. However, it does claim that children born into such an environment also enjoy the advan- tages of having parents and of being surrounded by peers with parents whose life achievements provide self- evident proof of their genetic superiority; in other words.I that social class Stratification is the product of a selection process that is. at its core. htogenertc and neturai To support such a vow obiectively. hard evidence would have to be produced of such genetic superiority or infev I'lDI'lf‘f as the determining factor of social class status, and what it means to be advantaged or disadvantaged generi- cally would need to be operationally defined. Moreover, children with low PPn would have to be proven to have parents who have :‘rgfefiorgenofipes, and a demonstra- tion would have to be made of why differences in the material environment alone cannot explain differences among social classes' performance on measures of PPA, reading achievement. school success. and later career choice. But Matthew EffecE offers little obiective evi— dence of this kind to support its etiological claims. In5tead. on close reading. the arguments of Matthew Effects appear to be based on a set of normative as- sumptions about the tightness of capitalism and about the natural order of the social order—arguments that his— torically could have been lifted almost directly from a Ich-centtlry treatise on social DarWir'i'tsrn (see Gould, lS'E-lfl‘Q'S‘fi'. Hofstatlter, 1944). Beginning to Read. a second publication that has significantly influenced recent cumculum policy making in Texas and California is Beginning to Readffidams. 199(1). It was commissioned by the Center for the Shady of Reading after the U.S. federal report, Becoming a Motion txfi'i'ena'ers. called for "phonics first and fast" (Pearson. 1990. p. v}. At 424 pages. it represents perhaps the most comprehensive attempt to review all the rele— vant research to date on PPA and cognitive models of early literacy acquisition. The first lCI chapters ofthe book deal mainly with reviews of previous research and begin with the statement that on the question of whether it is important or not to teach phonics, the conclusive evidence of fit} years of research has shown that reading programs that have emphasized phonetic decoding in some way have nearly always produced superior reading achit'vetnent in first graders. Beginning to Read declares The Great Debate resolved; the only questions left to discuss involve details such as how much, when, and in what forms phonics instruction should be presented to beginning readers. Chapters ti through Sr of Beginning to Read are a very detailed review of experiments] research on early literacy acquisition. The central task of these chapters is to build a model of the reading process that essentially reiterates and elaborates upon the model of literacy ac- quisition described in Starlet-rich (1935). Like Stanovich {I986} and juel et al. (1936} also. Beginning to Read ar- gues that prereaders who are most lilter to have trouble learning to read in school come from homes where they most liltely have no magnetic letters on their refrigerators, no home computer will! word and letter games, no reading clasrr tnatcs in preschool. and no ready supply of papers. pert- cils. and oayons lying around the house for their use. it is equally easy to imagine children whose television, if they have one. is preoccupied with programs that are less. directed to scholastic readiness of preschoolers than is ‘Scsarne Street." (p. at; As evidence in support of this fictional home, the ethno- graphic works of Shirley Brice Heath [1985) and William Teale (19136} are cited, and the comdusion is drawn that children who come from homes where storybook litera— cy is not a frequEni event are at a distinct disadvantage when it comes to acquiring literacy. Chapters ii] through 15 of Beginning to Read focus on the instructlonal implications of the first nine chapters of the book. A connectiont'st view of learning is ad- vanced that emphasizes a combination of systematic drill and practice in component subsltills with integrative lit- eracy acts focused on the active construction of meaning by a reader or writer. Yet these chapters also state at cer— 274 more ttttseaactt cosmetic? Jutyttugiovsepmmber 1999 W3 tain points that not all phonics or phonemic awareness programs are of equal utility or value. In Chapter 11, “On Teaching Phonics First." the issues that have tradi- tionally concerned the developers of phonics-first pro- grams are systematically described. and a discussion of the pros and cons of each conoem suggests that in each case the problems of a particular approach outweigh its benefits. After more than half the book argues for the teaching of phonics and for a model of word roosgnition that emphasizes phonological and orthographic process- ing as individual subcomponcnts of a mechanically oom~ plex system, the chapter states that "individual letter—sound correspondences and phonic generalizations are inherentiy intractable when divorced from the rest of the readirtg situation" (p. 291). This is because In the reading situation. as in any effective communica- tion situation, the message or text provides but one of the critical sources of information. The test must come from the readers' own prior knowledge. That knowledge does not function hierarchically from the bottom up but meets the printed information from all leveis at once, interact Lively and in parallel. Whedter to support learning or flu- cnt performance. it must. (p. 291) In the last 100 pages of Beginning to Read, a case is carefully rrtade for an early literacy curriculum that emphasims the interaction of the subcomponents of the model of word recognition described in earlier chapters, the reciprocal relation between the development of or- thctgraphic and phonological knowledge as well as be- tween the reading of eatended text (sentences) and individual word recognition, and finally, the regency of the reader in the constntction of meaning from a text. Noting that Phil is an ability that children cannot be told how to acquire, the possibility is discussed of using on- sers l{the initial consonant sounds)I and times [the closing vowel or vowel-consonant sounds}I in syllables in lin— guistic games and word play so that the "critical step of Lsolating and recognizing phonemes might ensue much more tractably" tip. 318}. To teach print awaremss with kinderganners. Beginning to Retard makes a strong case for the language Experience Approach (Allen. 19%}. in which children dictate a story to their teacher, who writes it on a chart exactly as the children tell it and engages their interest in the process of translating spoken language into text. The final chapters of the book also adsmate the use of in- sentedspelitng—encouraging students to use their bur- geoning orthographic knowledge to write words as they bEIiet-E they are spelled—as a way of funher developing their PPA: The Evidence that invented spelling activity simultaneous- ly develops phonemic awareness and promotes under- standing of the alphabetic principle is extremely promis- ing. especially in view of the dilfieulty with which chil- dren are found to acquire these insights through other methods ol'teaehing. (p. 38?) Moreover. there is no evidence produced that invented spelling confuses or retards the development of conven— tional spelling practices. On the contrary, a study try Clarke (1938} is cited. itt which students taught invented spelling and students taught traditional spelling both displayed curlsiderabie but comparable dilficulty in spelling highvfrequency but ord'tographicaily irregular words....However...the children in the invented spelling group were significantly more successful with both a list of iowcr—hequency regularly spelled werds and with the words on the Lovel ] Spelling Subtest of the Wide Range Achievement Test. lip. 334) In the concluding chapter of Beginning to Read “The Proper Place of Phonics." the role of phonics in— struction in the acquisition of literacy is reiterated, but the “negative attintdes" of teachers and students are also noted and sympathized with. since they can be “traced to the prevailing realities of instructional delivery" tip. 417). that is, the Stressful and tedious work of intensive instruction that is especially unrewarding and nonsensi— cal For children with low FPFL. However. “the problem is not really that children are given phonic instruction, but that they are given a relative surfeit of phonic instruction and in suboptimal ways" (p. 419). “Suboptiml’ ap- proaches are described as those that “enibodtyi the rnis— guided hypothesis that reading sitills are best developed from the bottom up" (p. 422}. or that proceEd from the assumption that there is "any particular. best method for teaching reading" [1]. 423}. Instead. diocis'tons about in- structional practices should be guided by “essential prin- ciples and goals and {clarified issues) that should be considered along the way" (p. 425}. Programs that are consonant with these goals, principles, and clarified is- sues are Reading Recovery (Clay, 19??) and “Don Holdaway's model approach to 'whole language instruc- tion" (p. 422). as mil as *Jeanne Chall's Reading Laboratory at Harvard {and} the program at the Benchmark School in Philadelphia“ tp. 421). Two points will connect Beginning to Reedto Matthew Effects and to the curriculum policy Statements of the next section. The first. which relates to both Stanovich (1936} and Juel et al. (1936}. has to do with the text's characterization of the literacy practices of low— SES and minority children's home communities. In con- trast to Stanovich and juel et al., however, Bflfflflfflg to Read offers actual ethnographic evidence by Heath (1985} and Teale [1936}. who studied the borne literacy practices of Appalachian. African American. and Mexican ‘I‘wostatee‘readingiuiflaitives 2T5 American working-class families in support of its claims. In these studies. both Head-i and Teale found that litera— cy was very much present as an activity in these families, but that there were telling differences among different ethnic groups, and bemeen these groups and white middle—class families. in their actual practices. Their analysis of their data attributed the problems that chil- dren from these homes encountered in school not to a lack of print or phonemes-“phonological awareness, but to an incongruence between assumptions made by the standard literacy cuniculums of schools and the chil— dren's borne literacy practices. a disiuncture that. they ara gued. teachers and schools did not see. Both Headt and Teale argued, in conclusion—and Heath demonstrated through her intervention in the curriculum of the schools her panicipant children attended—that once schools and teachers began to accommodate and build upon the home literacy practices of children (rather than insist that families change their ways with words}, the problems that ethnic, workingrclass children have in ac- quiring literacy could be dramatically reduced, if not vir- tually eliminated. In direct contradiction to the findings of both Heath (1%3) and Teale (1936} about the preschool lan- guage and literacy practices of working-class, ethnic- t'ninority families, Beginning to Read characterizes these children’s literacy experiences as being altogether too few, too short. and qualitatively irrelevant. also in direct contradiction to both Headt’s and Teale's conclusions that schools must begin to build on the strengths and skills that such children outside the “mainstream” (Heath, 1935. pp. 10—12) bring with them, Beginning to Read concludes that such students enter school with few if any linguistic or literate experiences relevant to achiev— ing literacy and must be instructed as though they were devoid of any prior experience in the analysis of spoken or wrinen language. Again, these conclusions place the arguments of Beginning to Read in opposition to the findings and conclusions of Heath's and Teale’s work: yet nowhere in the text is this opposition noted, not are differences in the interpretation of findings discussed. Instead. the arguments advanced by Beginning to Read are presented as though they are mere restatements of Heath's and ‘l'eale‘s own conclusions, Like Matthew Effects, Beginning to Read resists linking the problems it identifies with such homes direct- 1y to economics. history. or ethnicity. although unlike Matthew Effects no mention of parental genotypes is made. Instead, the claim is made that ethnographic research makes clr that poverty is not the major detemiinant of the lheracy preparation a chich re- ceives at home. There are homes that do not encourage preschuclcr‘s literacy skills in the {storybook literach ways described above. These homes are bust identified by neither in- corrne. social class, paternal education, nor race but by the values and styles oftlte social communities to which they belong. Childrtn from these homes not only miss the lit- eracy cuddling of their parents but grow up in a. larger environment when: reading and writing are peripheral and peripherally valued activities. lip. ii-T-‘l This passage begs the query: If the "values and styles" in question originate in external factors associated with "neither income, social class. parental education. nor race l(that is, ethnicity)," then what might be their source? Beginning to Read does not consider this ques- tion. The only answer available to readers who try to fol- low its reasoning closely is either that the values and styles of communities whose children struggle to read exist as their own first cause. or that their source is inter— nalin nature and located within the backgrounds of End dividuals and individual families. This must be the case. even drough these families are clearly not distributed randomly in the population, but are found clustered tow gether in conununities that Beginning to Rear: claims are only coincidentally marlted by clear ethnic and socioeco- nomic differences that only appear to separate them from communities where children routinely succeed in school. The exemplar of such a community offered by Beginning to Retard—a conununity where children “fall into a pattern of school failure” (Adams, 1990. p. 38}. but where ethnicity, social class, income. and parental edu- cation apparently piay no patterning role—is one de- scribed by Heath. and readily identifiable to readers familiar with Heath’s work as Trackton. However. in the description of Tracltton provided to readers in Beginning to Read Tracktcrn is not named. nor is the ethnicity, his— tory, social class, or educational background of any of its members ever mentioned, even though to readers of Way; was Words, the most salient and memorable de- tails of Heath's ethnography of this community are al- rnost certainly centered around the fact of its African American and working-dass heritage. This omission, in combination with other claims that are also contradicted by the full evideme of Heath’s and Teale's work, sug— gests Lhat the use of ethnographic evidence in Beginning to Read is not communicative in its purpose, but is in fact strategic. [rt other words, from a Habermasian perspective, it is oriented not toward the achievement of undemtanding with readers. but toward the achievement of an overtidr ing goal. That goal would seem to be to use ethnograph- ic data to highlight differences berween the home lives of children who stmggie to read and those who do not, are assume assesses qrrartrstttv Jutyrairgrevscpmmhu 1999 34:9 while avoiding the full implications ofthat dara's analysis for the reform of school literacy. For where the worlt of Heath and Teale demonstrates that some children suc- ceed in school because their homes and their sdrools share ccrtain normative, socio—. and etl'rnolinguistic as sumptions about how spoken and written language should work, the work of Beginning to Road is to demonstrate that the home literacy practices of success- Ful readers are not normativer but objectively better. ap* parently because they produce measurably higher measures of PPA and print awareness. Denying the salience of external structural variables such as social class. income. and ethnicity avoids the issue of social eq— uity in the reform ofsdaool literacy mat Heath's and Teale's work suggests must be addressed and redirects the discussion of reading reform to the instrumentalist question of how best to increase the PPA and print awareness of children whose families' values and styles fail to measure up. But as the analysis above also suggests. this argu- ment is convincing only to the extent that readers are unfamiliar with the work of Heath and Teale. or its con— tradictions with their work are overlooked. To skeptical readers with detailed prior knowledge of the history of The Great Debate and of its research traditions, Beginning to Read presents itself as a remarkably com- plex tear mat is open to multiple interpretations—that is, as an argument for phonics, but with many important caveats and qualifications. However, readers who were either not well read in qualitative and quantitative literaa cy research or not inclined to be skeptical about its ori- gins and its arguments might easily have a very different sort of reading of Beginning to Read. They might notice that in the opening analogy drawn bemmn reading and the operation ofa car, in which print is compared to fuel and the pans of the engine are compared to the sub- oomponents of reading, a driver is never mentioned. They might then conclude that reading is a reactive. au- tomatic response to a stimulus. When it is reported at the conclusion of Chapter 2 that "illiterate adults account for T5 percent of the unemployed. one-third of mothers receiving Aid to Families with Dependent Children. 35 percent of juveniles who appear in Court. (and) to per- cent of prison inmates..." (p. 2?), readers might easily mistake the correlation between literacy rates and social problems for a causal relationship. They might also con- clude that ifall the unemployed, the AFDC recipients, the juvenile delinquents. and the prisoners who are illit- erate could be taught to read, their problems would solve memselves. They could have read the approving remarks about Wyjohnrry Can it meme-sch, 1955, 1985) and the dramatic way the story ofjeanne Chall’s (196?) conver- sion to phonics is told in Chapter 5 and decided that a later critique of Flesch‘s politics and the chapter's con- cluding. vague references to qualitative differences among phonics approaches were not very significant is- sues. Sympathetic Iayreaders might be alternately imv pressed and confused by the thoroughness of the suboomponent-by-suboomponent argument in Chapters 5 to 10 for why efficient readers "process the individual letters of words" to. 103) but poor readers are slowed by context. They might then becotTre even more convinced that (a) phonics was a single thing and the only rational approach to reading instruction. (b) decoding was a process of sounding out a word by adding all its letters up, and {c} the teaching of conventional speJling and other tried-and-true methods were critical to early read- ers’ success. They would miss or dismiss these chapters‘ concluding remarks and qualifications of all of the above as insignificant deraiis. if they read as far as Chapter 11, "On Teaching Phonics First." they might take its title at face value. be confirmed in that assumption by its respectful discussion of traditional phonics-based reading programs, and miss the irony of its subtle discredit'rng of them on the grounds of their impractioility. misguided assumptions, and hidfin consequences in the final pages of that chapter. They might wonder at the approval of inset-trod spelling in Chapter 14 but choose to ignore its implica— tions; and if they read the concluding chapter of the book. "The Proper Place of Phonics." they could figure that the author is rnErely taking a magnanirnous. oolle- gial approach toward her potential critics. In short. it is possible to read Beginning to Read in good faith and conclude that the last it) years of reading research have been devoted to ironing out the details and filling in the missing gaps of what science knows about how reading works and what children need to have and do to acquire literacy. rather than read the book as the subtle and sometimes ironic discreditation of the programs of the 1963s. and a call for a major over- haul of the field's conventional wisdom not only about context as key to word recognition. but about bottom up, dccor‘rtextualized phonics approaches as well. But where is the evidence that this hypothetical scenario might coincide with some readings of this text? My evi- dence, ] admit, is circumstantial. This was my initial reading of Beginning to Read, especially of its first chap- ters: and it has also been the wading of the majority of undergraduates in a course [ teach where Beginning to Read is a required tend. This book has also had enor‘ mous influence on the reading initiatives of both Texas and California. Yet the complete lack of concern for the details and warnings embedded in Brigiflflfflg to Heads review of reading re5earch in the 30 years since The No states“ reading initiatives 27'? Great Debate. especially with regard to tlte role of phon- ics and spelling instruction. can only be explained either by the fact that officials in these states have not read the entire book, or that their reading of it was as cursory as the reading scenario] have described above. Decoding two states' reading initiatives There is, then, a clear discrepancy between the va- lidityr of objective truth claims made about phonemic! phonological awareness and its relation to reading achievement. and the validity of claims about the social. cultural. and pedagogical implications of PPA—claims supported largely by conjecture on the pan off-tarrhew Effects (Stanovich. 1936) and by the strategic use of ethnographic research evidence in Beginning spliced (Adams. 19%}. Yet there is very little oven mention of a relationship between PPA and the social. cultural, or economic conditions of children's lives in either state's document. The California document. Teaching Reading: A Balanced, Chmprebensr‘re approach to Teaching Reading in Prekindergarten through Grade Three (California Department of Education, 1996}, on the other hand. avoids referencing its claims to any population whatsoever (indeed, the children who illustrate the pages of one of its 1t‘i'e'eb sites are drawn faceless—al— though they do hold up letters that spell the word cowl. Screening for FPA uncch these guidelines is included with other screening procedures that are recommended for determining prereaders' print awareness and lan- guage proficiency in English and other home languages. like PPA research, this document never takes into ac- count how bilingualism, much less differences in English-language dialects, might afiect the results of those tests. How is the omission of any mention of cultural or social difference. in two documents that aim to address the needs of children in two states with populations as culturally and socially diverse as California and Texas, to he explained? One line of reasoning may be that they operated on the principle that the best way to achieve social equity is to construct a single policy and apply it to all students in a uniform manner; by this logic, the best way to enact this principle would be to treat socio- cultural differences with benign negleCt. n more prag— matic but perhaps less sympathetic argument sees the issue of sociocultural difference as simply too controver- sial politically to be accepted in an official state docu- ment. Yet as I have demonstrated in this study, and as 1Willis and Harris (1991'?) demonstrated in their revisitation of the Pitts Grade Studies of at] years ago. to tonignly or otherwise neglect issues of difference does not eliminate those issues or reduce their salience. Rather. it makes them present by their absence—that is, it leaves differ- ence, and researchers‘ and policy mal-ters' unexamined perceptions of difference. to operate in hidden ways that may lead ultimately to the invalidation of research find- ings, and often to policies with consequences quite the opposite of what even their best intentioned framers hoped to achieve. However, it should also be noted that even if the presence-hyvabsence of sociocultural difl'erence is dis— counted, the validity of the Texas and California docu— ments' primary claim that they are research based remains suspect. While on the surface both documents appear to take their guidance directly from the research. they do not discuss that research’s careful analysis of the limitations of PPA instruction and its relationship to word knowledge and the reading of extended texts. nor do they ever discuss differences among skills-based ap- proaches or the implications of connectionisr theories of reading, as do the texts analyzed in this study. For exam- ple. they include none of the analysis of the reciprocal relationship between phonemic awareness and word recognition in Matd'tew Effects, including the discussion of a ceiling effect for PPA training. Moreover, Beginning to Reads endorsement of phonics instruction carries with it clear caveats about the excesses and misguided as- sumptions of the past. includes a strong endorsement of invented spelling and the whole language program of Don l-loldaway. and notes that "knowledge l[of how to read) does not function hierarchically from the bottom up but meets the printed information from all levels at once. interactively and in parallel” (p. 291]. Yet the Califonua and Texas documents support the teaching of multiple subsltills in isolation {with some storybook reading thrown in on the side} and malre little or no mention of the research-supported innovations of the past 313 years, such as invented spelling and the use of onsets and rimes. in short, these documents appear to be using the term research to endorse views of reading as a process and its instnrctional implications that are grounded in the assumptions of the era of Rudolf Flesch—assurnptions that, as Beginning to Rental points out. Eli} years of research find misguided. Teams. In the Texas document, Beginning Reading Insomtion {Texas Education Agency. 1997.). tacit sup- port for traditional, as opposed to research-based, ap- proaches to reading is accomplished largely through the use of ambiguous language. For example, although “twelve essential components" of learning to read are listed "in an order that could imply a sequence ofin- sttuction” and are described as “interrelated” to. 2). little attention is paid to what connectionist models of reading suggest the nature of those interrelations might be. Unlike Matthew Effects and Beginning to Read, no men- tion is made of the reciprocal relationship octween the are assume assesncu QUARTERLY Juiymugustr'strptember 1999 airs development of phonemic awareness and word knowl- edge. or of the demonstrated value of whole language activities such as emergent writing (allowing preschool children to writer’draw stories and experiment with writ— ing before they've learned the alphabet). or language ex- perience approaches in developing word and print awareness in young children. Instead. the document calls for highly directive “activities” to promote discrete objectives such as segmenting words into sounds. "lessons in word awareness" (p. all, and lessons where children “learn the names of letters and learn to identify them. rapidly and accurately" (p. 5:]. The 12 cornponents are presented discretely and hierarchically, front lan- guage and print av-rareness, through decoding and spelling of regular words. to wider reading. and finally. to learning and applying comprehension strategies. The word phonics does not appear in Beginning Reading Instruction; but phonemes and letters are iden- tified as the "building bloclts” of written and spoken lan— guage. and "decoding instruction (that) is explicit and systematic" is described in very general terms (e.g., “practice activities that involve word Families and rhyming patterns"; "practice activities that involve 1olend- ing together the components of soundeth words” lp. Tl}. The text never refers to Beginning to Reads warn- ings about the misguidedness of bottom-up instruction and sounding out words letter by lener. or that many children do not transfer practice in decontextualized phonics activities to reading extended text without sub- stantial assistance and encouragement. Beginning Reading instruction demonstrates no awareness of the important benefits of increased phonemic awareness and deep orthographic knowledge drat Beginning to Read notes accrue from invented spelling, and instead states that "[als children progress. well organized. systematic lessons in spelling will be beneficial" to. Ti}. Coiifornio. The Texas policy statement appears a model of progressivisrtt when placed beside the first part of the California document, Teaching Reading {California Department of Education. 19%). however. For example, whereas the Texas document describes its essential com- ponents in terms of the opportunities they provide for children to act on and interact with printed text. the in- structional components specifically named in the California document seldom take note of the reader as an active participant in are reading process. but instead name the components themselves as the agents of its reading program. In its discussion of its second compo- nent. "Letter Names and Shapes." Teaching Reading tac- itly advocates an emphasis on penrnanship and against emergent writing, arguing (without referencing the claim)L that "until children can comfortably discriminate the shape of one letter from another. there is no point in teaching letter-sound pairings....[C]hildrcn will not write willingly until they can form the letters with adequate ease and to their own satisfaction" tan. The fourth component of the California document stresses the need for “systematic. explicit phonics” in— struction. and defines phonics as "an organith program where letter-sound cot-reapondences for letters and letter clusters are directly taught: blended; practiced in words. word lists. and word Families; and practiced initially in text with a high percentage of decodahle words linlted to the phonics lesson" (up). It repeats standard mg“- ments that proficient readers read every word; it cautions that "children must not be taught to skip new words or guess their meaning," and argues, with citations from ar- ticles published by researchers reviewed in this article. that ‘only poor and disabled readers rely on context for word identiEiCation” {op}. The document does contain a statement that "phonics instruction is not about rote drill involving a comprehensive list of spelling—sound corre- spondences and phonics rules" top}, but it never eit— plains what might be substituted in place of rote drill; instead. the sentences that follow and support this state- ment argue mat "the most effective phonics instruction is explicit—that is. taking care to clarify key points and principles for students. In addition. it is systematic—that is, it gradually builds from basic elements to more subtle and complex patterns" lI'.n.p.]-. Nowhere is the difference between "rote drill" and explicit. systematic instruction in “key points and principles" articulated; instead. readers are instructed that "lessons should focus on short words that adhere to the basic left-to-right principle of sound- ing and blending. such asli‘ai and fit'—-advice that ap- parently contradicts Adarns's {1991)}, Tunmer and Nesdale's (1935). and Lihemtan et ai.'s (19.74) warning against trying to blend the letters of a word together. In its Fifth component, the California document states dtat the use of temporary rpeiiirtg “lays a strong cognitive foundation for both formal spelling and phonics" but it also warns that “children should be expected to attend to the correctness of their spellings in their writing,” and that "children who scribble need support with print awareness and letter knowledge" (op). The term in- tsniedgoeiiirtg, which is commonly used in the research literature. is replaced by the invented term temporary seeding. although apparently the M terms refer to loon— tical activities. The last three components named by the California dooament-uvocabulary development. comprehension and higher order thinking. and appropriate instructional materials—are all described as ancillary to the develop- ment of automaticin in word recognition through the teaching of phonics. Vocabulary development is charac- terized as a function of the ability to decode effortlessly. Two states“ reading initiatives 275‘ The discussion of comprehension does little more than restate the argument for autnrrtaticity and. implies that if children can decode and recognize words with enough speed, comprehension follows as an automatic by-prod- uct. Finally, although the section of the document that describes appropriate instructional materials mentions that a "balanced. comprehensive early literacy program must embrace a variety of reading materials" (up), the next serve-n paragraphs of the section argue that early readers must only read deoodahle texts that are com- posed of words that "should be wholly deoodahle on the basis of phonics" {although they can be read to from other texts; n.p.}. Much of the second part of Teacbe Reading, "Graderlevel Expectations and Examples of Classroom Practices," does not follow at all from the first part. Learning activities approved as prekindergarten. kinder- ganen, and first-grade practices in this second pan only indirectly address explicit instruction and never mention the use of any sort of Systematic phonics program. However. California Assembly Bill Hit-i6 (199?). which appropriates funding for inservice teacher training in the new reading program. specifically identifies 12 discrete suhskills that will be the subjects of this inseniee. This bill makes no provision for the implementation ofthese suhskilis within the sort of functional. communicative ac— tivities described in the "Expectatioro and Examples" part of Teaching Reading. instead. it is careful to state that “lslystematic explicit phonics instruction hruilds from ba- sic elements to complex patterns and teachers provide prompt and explicit feedback. Systematic explicit phonics instruction does not mean ‘etnhedded phonics instruc- tion’ which is ad hoc instruction in phonics based on a random selection of sound and word elements” (up). Stimulus-y and discussion: Two statesl' reading initia- tives as strangle, not icor:t:ti:t‘.tt.tnicat:tttl'e+ action in summary, recent reading initiatives in two states have based their curricular actions on the claims that there is now sufficient scientific evidence to state conclu- sively what is l{and by irnplication is not} sound instruc- tions! practice in early literacy education, that the cornerstone of this evidence is phonemic-“phonological awareness research; and that the future of children as lit- erate individuals in these states is ieopard'tzed by current cun-iculums based principally on the reading and writing of meaningful LeJtts. These initiatives also hold that those cun'iculums need to be replaced by explicit, systemch instruction in suhskills such as PPA and pl'ifll'iitB that are backed by assessments that will screen young children for apparent linguistic deficiencies with the goal of linemen- ing in their language development at the earliest possible moment. The task oftth article has been to investigate the relation of curriculum policy documents in these grates to research reports and reviews of pl'iot'tvei‘t'lit.'..-'r phonological awareness. Using a framework suggested by Jurgen Haber-mass Theory of Conununicative Action (198?; see also Carspecken. 1996). l separated research reports. with validating criteria that are objectiver refer- enced, from reviews ofthre research, which make norma— tive-evaluative use of objectively validated facts. and from curriculum policy statements that rely largely on research reviews to formulate pragmatic goals about what is best practice within the current social environment. A close reading of these curriculum policy state- ments and. their supporting evidence shows their propo- sitions to be not so much research based as research traced. Reading stern, l was unable to detect any sign of a thorough review of literacy research of the past 3i) years that could be named as the foundation, or base, of the policies named in them. They disregard or avoid— and in the case of California’s position on contextualized phonics approaches and invented spelling, actively and directly conttadict—any research findings of the inter- vening 50 years that refute or adiust traditional beliefs about the efficacy of traditional practices. Rather than ad- dress the research—supponod call for tasks that would encourage early readers to integrate and coordinate their developing knowledge of how reading and writing work as systems for producing meaning, these programs pay little if any attention to the cormnunicative function of language, to the complex interrelations of the subcom- ponents of literacy. or to the agency of children as loam— ers or of adults as teachers. Instead. they prescribe programs of explicit, systematic instruction in the tradi- tional subcomponents of reading and writing plus pre— requisite screening and intervention via phonemic! phonological awamness activities. Moreover, rather than take any note, either tacitly or overtly, of objectively validated research of the past it} years that demonstrates the salience of social and cul— tural difierence in children's acquisition of school litera- cy. like the Fiat Grade Studies of 196? they tacitly maintain that literacy is a cognitive issue that demands an almost exclusively cognitive approach to its instruc- tion. In short, rather than read like documents construct- ed on new foundations laid on new terrain, the reading initiatives of Texas and California read like remodeled lists of standard, skills-based methods and procedures that have been updated by the addition of a few new components, such as phonemicr‘phonological awareness, and a few new phrases. such as higher order thinking. M a consequence. there is no evidence in these documents that would discourage a teacher from decids ing that a «in or 5-year-old from a low-SE5 or linguistic minority background. who either was confused by his or ass READING Rssnancu We Jutyrsugmtrseptember 1999 54:3 her teacher's accent or by the strangeness of PM. assess- ment practices, was linguistically deficient and headed for academic disaster without a quick, intense dose of FPA training and phonics. There is also very little in these curriculums' notions of balance to discourage an early literacy teacher From spending the maiority of in— scructional time teaching phonemic awareness and phon- ics tasks from a purchased program, giving weekly spelling tests. using the extended texts of highly con. trolled basal readers only as devices to practice and develop automaticiw. and reading a story aloud occa- sionally to break the monotonous labor of learning to read as an unnatural act. Similarly, there is also little in these documents to encourage innovation in any form, from teachers developing their own phonemic aware- ness and phonics lessons based on stories and poems. to the development of phonemic and orthographic aware- ness through invented spelling, to the teaching of multi— ple strategies of word recognition and oomprehension in coordination with tests that address subiect area knowl- edge in ways that engage children's intellects. It is not onlyr a particular type of research or nostal— gia for the 194565 d'iat motivates diese policy documents' calls for a realm to skills-based instruction in reading and writing, but an apparent need to eliminate any trace of the language and beliefs of the policies that preceded them as well. in California. this is Evidenoed within Teaching Reverting by the substitution of the term tempo- raryspei't'mg for the conventionally used interim!“ mefi'ing, by the avoidance of the terms firemtatre and hit- owy. and by the clarification that embedded phonics is ad hoc and not a suitable substitute for explicit, system- atic instmction. it is also indicated by provisions in California Assembly Bill 1036 (199?) that directly prohibit any person applying to give an inservicc presentation in the state who uses the terms incentive (sic) speiiirtg and Embedded phonics in their presentation from being paid with state funds. In the Texas document, a subtler avoid- ance of previous policy language is evident in the ab— senoe of any mention of invented spelling, emergent literacy, ettpressive writing activities, and word play, as well as in the repeated, heavy emphasis throughout the ten on 'opport'unities" to "learn" and “practice” “sound- letter relationships,” "effective sounding-out strategies,“ and "direct instniction in decoding” (pp. 6-3; see also Helfand, 1993)- The reasons given by the officials of bodt states for making such dramatic and controversial shifts in their lit- eracy policies are also problematic. In California, poor scores on the 1992 and 19514 National Assessment of Educational Progress (NAEP) acre blamad on the llnplEr mentation. in 198?, of policies advocating a more holistic approach to literacy instructim‘i. its McQuillan (19931 notedI in rank comparison, California was in the bottom third of states in 1992 and in the bottom fourth in 1994. and so. based on this data alone. there did appear to be cause for alarm. Yet McQuillan also pointed out that “performing poorly is not the same as getting Horse [ital- ics in original!" (p. 13). To establish a connection be- tween the policies of 1937 and the NAEP results of 1992 and 1994. it would have to be shown that NAEP scores in California were higher prior to 198? than they were in the 1996s. But this connection cannot be demonstrated. because state—level wasp scores are unavailable before 1992. According to McQulllan, the only test scores availr able that indicate levels of reading achievement from be- fore and afler 198’? are the California Achievement Program scores. whose most salient feature is their sta— bility from 1934 to 1990. In fact. the raw soores for this test show a slight [although probably not statistically sig- nificant} gain for 5rd, 6th. 3th, and 12Lh gradem from 1984 to 19911 (p. 13:}. McQuillan concluded that while reading scores in California woe not high in comparison to other states, there is no indication that the policies of 198? had any dismonstrable died at all on the state's reading achieve— merit scores. Other observers of reading achievement in California have pointed to the number of boolts per student in the state (at the elementary level, 13:1 in California, compared to 13:1 nationally}. and per pupil ii- btai'y expenditure {at the elementary level. US$3.48 for Califomia, compared to $15.44 nationally) as its possible cause (Krashen. 1996. p. 41. it also seems relevant to note here that in 1995, when the Losidntgei'es Times stat- ed on its front page that California's reading scores were ranlced behind Louisiana's, it quoted a teacher on page 24 who differed with that conclusion. pointing out that “[tlhe problem is we have 2.9 lcids in a classroom, 60% of the students don't speak English, and we don't have the resources to give them the support they need" (Colt-tin. 1995, p. n24). Almost entirely absent from me media's analysis was a correlation between Califomia‘s test re- sults and its teacher-student ratio of 24:1, compared to the national average of 11:1. Or an average per—pupil ex- penditure of 543%, which was well below die national per-pupil rate of 55.594 {cited in Dressrnan, McCarty, 3; Benson, 1998). in Texas. the pressure for reform is bout executive and media driven. The Houston Read Commission, for example. widely disrributes materials citing results of the Texas adult literacy Survey claiming that 52% of all adults in the state are fianctionally illiterate (Houston Read Commission. 1998). These reports are accompanied by newspaper articles that feature graduates of local uni— versities' teacher education programs complaining they were never taught how to teach reading—that is, never No stam‘ mdlng initiatives 251 taught how to us: a sldllsbased, published program of instruction like DISTAR (Maddey. 1993). They promote the claim that Texas is currently suffering a literacy crisis. brought on by an absence of instruction in the basics of literacy in its elementary schools and colleges of educa- tion. In the gubernatorial election year of 199-3. public of~ licials in the state frequently refer to Texas's reading initiative as the Governor's initiative. Backed by literacy action groups. his mother. and the media. Governor George W. Bush alludes to plans dun'ng his second term in office to "retrain" the state‘s public school teachers, ap- parently without the input or assistance of colleges of ed- ucation in the state. since that is where they were “poorly trained" in the first place by faculties caught in a “philo- sophical quagmire" about the best way to teach reading ("Bush Sketches His Vision for 2nd Tenn,” 1998, p. J6}. in Texas as in California. statistical support for the claim that tl'te state is in a literacy crisis is also question— able. Governor Bush has cited the results of the 199‘? Texas Assessment of Academe Skills {TN} to argue for a crisis on tl'te grounds that “lllast year, nearly 4fl,flfll} stu- dents failed the third-grade reading exam” ("Bush Sketd'ies His Vision for 2nd Term,‘ 1993, p. Id). That would seem a shocking statistic. one implying a crisis within the state's teacher preparation programs that dev manded an immediate and radical change. Yet there is no data available that would show what percentage of those autos students were taught by teachers actually educated within the state. Nor are figures available that show what percentage of those 40,000 students. in a state with an extreme teacher shortage and where as many as balfthe teachers in major urban districts are hired with emergency ceniflcation (i.e., who have a bachelor’s degree and are enrolled in an accredited teacher preparation program} were even taught by fully prepared and fully certified elementary teachers. a case in point is the postbaocalaureate student working toward alternative cenification in my sopho- more—level course at the University of Houston. entitled "the Reading Process. she is teaching first grade this year under emergency cenification. was given her current as- signment after one year's experience teadting second grade, and has had no instruction in reading methods save what she could glean from the teacher‘s manual of her school's basal series. And although it is conunon knowledge among bureaucrats and teacher educators in the state that many rural and less affluent school dis- tricts—districts with the highest number of at-risk stu- dents—hire individuals with 2 or 3: years of college as long-term subs {and so skin the law that teachers must hold a bachelor's degree), there has been no accounting made to date of how many such teachers have been hired, what their assignments are, or how many of their students passed the third—grade Tm in reading each year. Finally, even when these issues are taken into con- sideration. a more longitudinal analysis of rats results (see Table 2) shows steady gains in third-grade reading achievement from 1994 to 1993 in the state, with the highest gains made by p-Opulations most often consid- ered at risk: African Americans. Hispanics. and the eco- nomically disadvantaged. By the logic ofthis claim, were the present average yearly rate oi" improvement for these groups of 2.8%, 2.15%, and 2.6%, respectively, to continue for another 5 years. the litemcy crisis Govemor Bush a1— ludes to in campaign interviews would apparently re- solve itself and so would appear not to be much of a crisis at all. Whenrhe clairnsofaliteracyaisisinthesetwo states are examined in light of the selective use of statis- tics by politicians and the media in those states. and in California by legislation, the evidence strongly suggests that the use of research and scientific evidence in these documents. from a Habertnasian perspective, is strate- gic—that is, goal oriented—rather than communicative, or oriented toward bringing reading educators and administrators in schools and universities to an under— Tahle 2 Texas Assessment oi" Academic Skills crass) Grade 3 statewide results for reading Percent passing with rust or better correct multiple-droid: responses Demograch group Tfiflng year Yearly gain in percent 19'“ 1995 I995 199'? 1995 9-5—95 95-96 95-9? SIT—9‘3 94-98 Nuispecial education TN: 79% arm 31% 86% 2 I 1 5 9 African American 62% 65% 65% I39“: 76% ,5 I J- 'l' 14 Hispanic bill-ii m 7-‘2‘113 rats 31% 2 2 1 3 13 White sate iii-"is ssu eats are 1 t 1 5 I5 Economically disadvantaged 65% 68% T-‘D‘i'h 11% m 3 2 2. 'l' 13 Mair F'an Texas Education Agency tTEii} web site. .- -. -. . . as: READING REsEaRca QUARTERLY JulyrrsugusUscptcmhcr 1999 salts standing about the research findings of PM. and its im— plications for their own practice. A full analysis of What those goals and their implications for public education in the United States might be is far more complex than can be examined Fully in this study. However, the stategic. as opposed to communicative. quality of these two states‘ policies and implementation procedures becomes appar— ent when the linguistic and socioeconomic focus of the nonnative-evaluative assumptions inherent in the re- search cited in their support is considered in combina— tion with other recent analyses ofthe social, economic. and political climate of educational policy making within the United States (see Berliner, 199T; firessman, McCarty, 3r Benson, 1998', Taxel, 1997; Taylor, 1993}. Finally, when the effect of increasing immigrant populations and rapidly shifting demographic profiles on the school-age populations in most- two states are also taken into con— sideration (1.1.5. Bureau of the Census. 1995}. then the same implicit question raised by Willis and Harris (199'?) becomes salient once again. Does it make more sense to understand the anxiety of these two states about literacy rates as symptomatic of doubts about whether instruo tlonal methods will keep the U.S. globally competitive, or are they better understood as antiopatory of immi— nent, maior shifts in the internal sociocultural, economic, and political dynamirs ofthese two states and the nation? its in a revisionist reading of The First Grade Studies of I96? (Willis 3:. Harris. 199?}. to support the lat— ter daim above as more reasonable than the former, one need not claim that a conspiracy is at hand (at least among researchers}. but only that the research support— ing these initiatives has largely failed to acknowledge or examine its own normative assumptions about language, literacy, and schooling as social, cultural. economic, and political phenomena. If this is the case. then. as in Willis and Harris's reconsideration ofthe First Grade Studies, it is also pertinent to consider what the responsibility of re— searchers ought to be with regard to the ways research is conducted. conclusions are drawn. speculations about its educational significance are made, and how, in the end, it may be used to support curricular policies. Because from a [-laberrnasian perspective one can argue that it is not science or research per so that appear to be the problem. but rather the uses and misuse-s to which they have been put strategically, it may also seem that science in general and the findings of PPA research in particular are no more than convenient pawns in a political game in which curriculum policy and education— al research are more aptly characterized as il'tEUUIl'IEfltS than as perpetrators Such a conclusion might be a relief to literacy researchers. who could then argue that they have little control over the uses to which their work is put, and to policy makers. who could argue that the commexhies of the research findings have been misrep— resented to them. But the analysis presented in this study shows that a reliance on science as a general metaphor and on PPA research in panicular is far more than a rhetorical ploy. This reliance. in fact. is the mechanism that legitimates—that makes reasonable—a political process bent on controlling, not ini'onnlng, what univer- sity faculty are allowed to say about literacy instnaction. and how teachers can respond to the extraordinarily complex needs of individual students and classrooms. Science and literacy researchers, as well as educa- tors, are more deeply implicated in the potential inius tices these curriculum policies enable than they may either consciously realize or choose to admit, because. from a Haberrnasian perspective, it is their speech acts that legitimate the power plays of others who depend on public spheres of discourse. such as journals and confer- ence symposia, not just for cover, but for reason. Yet this is to the advantage. not the blame, of researchers and educators who strive to improve the literacy. and so the lives and work, of early readers, their teachers, and soci— ety as a whole; as reasoned discourse from a Haberlnasian perspective, provides the communicative means by which math claims can be sorted out and what is right can be distinguished from what seems expedient. I hope this study, along with other recent analyses of and responses to policy initiatives in Texas and California (cg. Krashcn. 1996-. McQuillan, 1993: Taylor. 1993), conducted by researchers with a far better grasp of the political dynamics of theae states and the research evidence with regard to literacy instruction than my own, makes some contribution to that sorting effort. For its part, I hope this study suggests one conceptual framev work, or common ground, for litenacy rescarchem of all methodological orientations and normative perspectives to begin to communicate with one another. For that process to move forward, I suggest that the conversation must move along two paths. Along one. the power of science as a metaphor and as a mediod ofdiscourse within the public sphere must be grasped and accepted by all parties. while what it means to sloscientitic research ought to be subiected to intense scrutiny and public debate. In the end, i think we could agree that science cannot only be a matter of counting and compang differences between a treatment group and a control group, for by that methodolong reasoning, Einstein would not be a scientist. nor would Darwin, nor would most astronomers and many biolo- gists. Instead, literacy researchers and the public must come to understand science as an activity that is funda— mentally and ruthlesst skeptical in its approach to the investigation of all observed phenomena, be immediate; ly wary of all conjecntre about what has not or cannot five states’ reading initiatives 233 be directly obsenred. and. in the case of human social behavior. be open to full investigation of, and consmnfly vigilant with regard to the implications of, its normative assumptions and biases. Such a stance would push the logical deficiencies of morali’biogenetic explanations for illiteracy and its consequences for society and the lives of individuals into bold discursive relief and raise the awareness of WA re- searchers about the sociofethnolinguistic complexities of measuring individuals capacities to identify phonemes. Moreover, if extended to all forms of empirical research and argumentation, it would also provide grounds for in— vestigation of the normative, historical, socioeconomic, and cultural biases of more holistic, expressivist ap- proaches to literacy acquisition as well (for a preliminary exploration of some of these biases, see Dressrnan, 1993:, 199?; Lensrnire, 1994', Moiard'tey, 1994; McConniek, 1996}. Along the other pathr critics might begin to chal— lenge within public settings what it means to read and write, to argue that these are: means to a conununicative, understanding-oriented end, and to reason publicly that communication, ifwe take Habermas as our guide, cle- mands that the agency of all conununicants must be re- specred. In sopport of this position, one can argue with force and integrity from objectively Tvalidated evidence that much of the PPA research that touches instructional issues {e.g., Bradley & Bryant. 1935: Cunningham, 199d; Vellutino & Scanlon, 198?; see also Adams, 199m has found drat PPA depends less on the capacity to react than the capacity to actively reason and use language as a system. In that case. rut-5F"ECt for the agentsr of nascent readers as well as the professionalism of teachers— rather than the claim that the urgency of a problem and its ends justify its means—stands at the ethical center of any literacy policy or curriculum. And teaching young. readers about how they might learn to use words in coming to know the world on their own terms, rather than to react to systems retrieved from their control, stands as policy making’s uncontradicted, communica- tive goal. Making the distinction between reaction and action consensually understood—and the difference that dis- tinction makes in the education ofratiot'tal. reasoning cit- izens and in furthering rational societal goals—is in the end the key to challenging current trends in curriculum policy making. For that distinction draws into sharp con trast the crude instrumentalism of the cunicular policies of California and Texas, where using and misusing re- search evidence. on close reading, produce little more than predetermined political. not educational, ends. “I'm inflamed Ten Phenolic-“phlan m tees} studies BMDIE'I'. l... 3.- BRYMI'I‘. P.E. {1935]. Categorize-Lg sounds and leamlng to read—a causal sonnet-rim. Nature. are. 419-421. CUNNINGHAM, A.E. (1990). Explicit versus implicit instruction in phonemic awareness. flattmcd offiwl‘msnml Child mm. 545'. 42944-4. EHHI, LC. {193?1 hunting in read and spell words. journal of Reading Mam-l“. 19. 5-41. mm. C. {1983). Leaming to read and write: it longitudinal study of 54 children li-om first through fourth grades. four-err! efficient-tritium Fathering}; so. {ST—44?. J'UEL c.. oaIFFrrt-I. P.L.. a GDUGH. Pa. (1955). Acquisition of literacy: .ir longitudinal study otf dtildrcn in first and second grade. firm-rm! qFEd'ucrarionral' Parenting}; PE, 243—255. tracer-tan. l.'r'., sI-Luttrwotea, o. FISCHER, am. e Cams. B. (1911). Reading and the awareness of linguistic segments. finer-rail gapefiarentai Child fissures-Ilng Id. 201-211. PERFEI‘I'I. CA. BECK. L. BELL. I... It HUGHES. C (195?). Phonemic knowledge and learning Io read an: reciprocal: 1L long't'udi- ml study of first grade el'tlldren. Mern‘l'i—Pnimer Quarter-tr. 33. 235—519. TUNMER. WE. & NESDALE. MI. (1935]. Phonemic segmentation sltlll and beginning treading. momentum mutant. W. 41 i—«izi. ‘v'ELLL'I'tNfl. Eli. & SCnNLDN. EM (196?). Phonological coding, phonologlcal awareness. and reading ability: Bidder-re: from a longitu- dinal and experimental study. Mfilfierfltmflarht ,13, 321—36}. YDPP, HIE. (1933). The validity and reliability of phonemle aware- ness lists. Heading Essence Quamefiy, .23, 159-177. Two lime reviews otrPI'rtt Inna-ch ADAMS. ill-1.] (199$. Beginning to read: Thinking and hunting abouiprim'. Cambridge, Mir: MIT Press. STANDWCl-I. KE. HEIRS}. Lhtfl'tew Edens in reading: Some core:- quences of individual dit‘ierences In the acquisition ol' liteficy. Reta-ding Resmrcb When}. 2:, inn—+06. Tens mt! censor-rd: reading inldflvea GiltFoaMa pessimism or EDUCATION. {1996}. nurses-.3 mailing.- A notarized. con-preseason approach to rears-reg reading or premises fer-neg}: grads titres. Sacramento. Cal: humor. Texas Eoucanott AGENCY. (1997). Beater-ting teasing tester:- Hoti‘lr Comets andfial'ures qfe march-lured reading program. Austin. TX: Author. Dflaer clwdworkrt All-EN. RV. {15%}. Entanglement-eaten in mater-titration. Boston: l-Io‘ughton Mil'l'lin. smarts, a. {1956}. Headlng acquisition ol' prerchnol children with- out systematlc lnsa-udnn. Eddy dammed: firm 1', 69—35. AU. EH. (1993}. Social cmme and the std-tool literacy learn- ing, ofsmdettts of dh'etse backgrounds. journal qfflterncy Mares, 30, 293L319. AUSTIN. J.L (1W5). How to do things wile nor-d5. Cambrldge, MA: Harvard University Press. mass 13.911993. Felzmaary 2). How good isithtsgood as it gets. governors say. The New Hart Tim. A12. BARREL RB. [1992}. The cultural gap in literature-based literacy instruction. Education and Urban Society, 34, 221—215. BARTDLGME L. (19941 Beyond the methods Fetish: Toward a ho- rnartizirig pedagogy. Howard Ednsflflonfil'iilfldnst 15!. 175—191. m READING RESEARCH QUARTERLY juhrrnugusrmcpremher 1999 we BATESH. _T. {1933-}. Black sir-eel gnocch- l'ls binary. mic-allure. and neural. ALnitii-L, TX- Universlqr 1:mee Preset. BEREI‘IEII. r1. s. FNGELW. s. [1966}. Teaching disadranraged children in rorpwocbooe Englewood Cllfi's. NJ: Prendoe-Hall. BEHLINER. D. (199?). Educational psychology meets the Chrlstlan RigJ-u: Dit'lering views of children. schocfling. reaching. and learning. Teachers College Record: 55'. 531-416. 13mm. R. (ISISIT. October so). Ecornmic oomcbodr: Housman works; ofl'loe space, h-IJ'LISE going fast as real mate comes back. Herero-r Chronicle (Horn-recs; p. 1. BUG-Go. 5.T. {1935). Searching. refining. and lemming.- A Mgr Hmom'i'an children a: home ands: school. Nome-ed. NJ: Ablcx. BOND. G.. & DYKSTM R. (1961319913. The mopcrnlivc research program in first-grade readqu hmmlon. Rmrflngflmmb Queries-fl; 3.3. 343-421 BUSH SKEl'CI-IES HIS ‘r’lSICIN FOR 2ND TERM. ("1993. September so). Ausn'rr Amer-iciammrmn. pp. _|1. Jo. CAIJPDENIA ASSEMBLY BILL 1055. {199?}. IDI‘Ilim'l Available: W CARSP‘ECIEEN. F.F. (1996). Critical slow in adamn'ormi re- search: A Theoretical andpmcrl'cal guide. New Yorlr: Eoutlcrlgc. mom. C. (1983). Classroom discourse. Portsmoutl'l, NH: Heinemann. CHAR. LS. (1963") “drums to mod. Mgr-ear debate. New York: Mcan-Hill. til-111.13,. 1.5. {1933) reaming ro read.- Tliegrmr driwefllpdaied EclJ. New York: Mcfiraw-Hill. error-ism. N.. s: HAILE. N. {19681. Tire mndpol'l'em qrsngnse. New York: Harper & llnw. CHRISTIE]. {1997. August 53'. Will new mmlurn starrinrds help our se'l'sor-lchilelreni-I Ber raised for Texas diIJdi-en. and Mil meet The challenge [Editorial opinionl. Harman Chou-ride. p. l (Dude-0k section). CLARKE. 1.11. {1933). Encouraging invented spelllng in first graden' wrifing: Elfocrs on teaming Io spell and read. Emir in die Teaching nfEnglim 22. 231—5419. CULY. MM. (1979}. Theer detection qr'rewdlng almonds. Porrsrrpmrli, NH: Heine-mam. COD-“IN. Ere. {1996. September 15). State repon urges room to basics in teaching reading. LosArrgslss flirt-es. pp. M. A24. BREED-om. M. (1995). Limos'mg lone 1involves: The cultural roman- licj of Literac].r workshops. Chmhrm Inquiry. 23. 245—165. DRESSMAN, M. (1997}. film in theiilbmrje Wing rbe graces Mums: orderandderira Weapon, CT: Bergln & Gan-re}: DRESSMN. M. MCCARW. L. .E BENSON. J. {1998}. Whole lan- guage 15 slgnifier. Conslriering the sonanfic field of school literacy. forum-r1! gore-cry lessee-ch. 30. 9-52- DTSGN. A. {19931 microfilm-Ms rafcbrlrr'ren learning R1 urn-ire in an urhanprmrjrsobml New York: Teachers {hllege Press. FLEGE. _r.E.. it FLETCHER. K. (1992). Talker and listener ell'ecrs on. degree of perceived accencJ'orrmalq’rlrsAcousriealjoclsn' or Ari-review. 93. STU-539. F1.EGE.J.E.. 51 PORT. ll. {1961). Cross-language plum-.13: interfer- ence: Arabic to English. Mrrgurlgeandm. 24. 125-146. FLESCl-l. R. (1955). Wbyjobnnymn'l! mod. New York: Harper a. flour. FLESCH. F. (1965). Woyjobnnycnn'r readflnd ed). New York: Harper & Rom. oEIoEchr-I. HJ. {1992) Englflbpbmflogc Arr [Hoods-radon. Cambridge. England: Cam'twidge UHMISHY Press GILLJCAN. C. (19321. In n rl'ajflh-errr puree mineralogical terror}- and worm-it's deeolqomenr. Cambridge, MA: Harvard Unioersiqr Press. 600W. K. {1996}. OH reading. Ponsniouth. NH: Heinernann. GCN'JD‘I'. _T. {19”}. The mammalian qfrlie Lounge mind. New 1for-Ir: Cambridge Univeiain' Press. GDUID. SJ. (RENEW). Tbs mismre earn-um. New York: “KW. Norma-i. GRAFP, l-I._[. (1981"). The refills-me}!- Cmflnrrlrlrrs and con- rm'i'cl'i'om in western sullen-Band soda-w. Elooum'nglon, IN: Indiana UniverSltr Press. CRAFT. HJ. {1991}. fire Mama-I myth.- merary antisocial mature in rlrs mime-cemrrw my. Brurlswiclr. NJ: Tommon. GRAFF. HJ. {1995). moowsnrrgmm Piusburgh. FA: Un'n'erslty of Pittsburgh Press. HABER‘i-LAS. _T. (193?). Hie rbeoryq’conmunlmnwacrmn, lrbi'. l.- imrson cadres Wort qfsoclnp (T. McCanl'iy. Trans). Bosron: Bacon HABEHMAS. _l. (1993). Modernity: An incomplete prolocl. In T. Mlfif {Eli}, Fosmodemim: A readeripp. 95—1139). New 1Erode: Columbia Urdverslnr Press. HEATH. 5.3. (1933}. Ways one nor-air. Cambridge. England; Cambridge l.l:'i.ire-r.-1ir5I Press. HEIAND. L. (1993. Climber 2'5). Reading by 9: Sane professors resist rule‘s reform formula. LosAnge-ln‘ fill-nor. pp. 1. 34-35. HDPSI'ADTER. R. {I 944}. Social Domenic": in American thought. Eamon; Beacon. HoHN. we. .9 EI-llll, 1.1:. [1985}. Do alphabet letters help pre- readers aoq'uire pinion-c segmermxlm slnilEI more! whit-ream! Psi-clam. 3'5. TEE—T62. HURKHHMEH. Iii-1.. it sooeno. T. (1995). mm arm-ringer.”- rns'nl' (j. Cumniirtg. Tram}. New Ym'lc: Confinuum. IUClr'LgLrtall}.r pub- Ital-Led, 1944) HCILETDN READWIRSIW: mull-CY NEWS. [1998}- lfirflinel. Available: hug-flmhgyufid 9mm I].|'.|II1 moons. S.L Elwilune 21]. Eucllen jus: wams to learn to read. LosAflgelcs Hines. pp. B5. 365. JENCKS. C. (1W2). Inequality Now York: Basic Books. JENSEN. LT. (1995). WNW. mm and Philariuiplua. John ERASHEN. 511996). Ever-pm a reader: Arr mosaic the Calyme Task Harte Equal-r on Reading. Culver Clry. CA: language Eduulion .rlsseclates. lei-BOT. W. (191'23. language in tire inner city.- Srtrrfles in fire black English merrier. Philadelphia: University of Pcnnsylvania Press. LASS. R. (1W6). Engflslrpbormlugrandpbonologlml recur}: Synchrom'c and diachronic undies. Chrnbndgc, England: Cambridge line-Hersh}.I Press. lliNSMll-lE. TJ. (1995:}. Who children iii-fire Cfiflcnl rs-rrrrrorrs of rise rev-lung mew New York. Teachers College Press. ICING. M. (1990). Mamrarimal consul-aim on language dovelopa fl'iE-‘fl'li. Err-mesh SecondbnngniagsAc-qrrlrflm 12'. 251435 HARRIET. M. (199?. October 11}. Bush dedicates UTs reading re- search unit: Illiteracy called a 1131»: childhood disease.‘ muslin: Commie. p. n30. MARKER. M. (1996. March 1). Battling illiteracy: Public school mailing instmctimt gelling a push from parents. policy makers. Housman Chronicle. pp. A1. A21. MDCAElTl-IEY. SJ. (1994}. Authors. ton. and tallr: The nonmalig- ticrn of dialogue from social interaction during writLrtg. Marlo-{g more}: gnaw. 2.9. zoo-251. MDCDRB-flCK. IL {1996). Tbs'cul'rum qfrmding audios Macho-g of English. Manchester. England: Manchester University Press. MCDEEMUTT. R. (mo. Athin school failure: An lnlhropo- logical approach lo illilcmqr and. social stratification. In G. Spindlecr (EdJ. Mondrian one! mwomceiswp. 52—113). New York: Holt. Rinehalrr. & Wharton. HCQUlLLuW. J. [1993}. Remover-Isis- Mocking. res-isom- Two steteE‘ reading initiatives 235 Hons. Portsmoum. N'H: Heinemsnn. MEEl-Ltlvl. M._T.. & WILLBRAND, 1.1.. {1919}. Lilde in children A resource boohl'br memnguagepaibolegisis. Springfidci. IL: Charles (2. 111011115 MELMED. PJ. 09?”. BM Englflbpmeiogy; Theme ty'" reading inieu’ererroefh'lonognph No. I). Blakeley: University of California Language-Behavior Research lab-oratory. MDLL, L {1992}. Literacy rfiearch in eomrnun'ly and class-Dorris: A mlmrfl approach. in It. Beech. M. Green. ML Kamil. & T. Shanahan (Eds). .Wiseieiimnlpemm or: fires-rig- res-scarce (pp. 211—244}. Urbnna. 11.: National Conference on RESCIH'J'I. in English. DISGN. D. (191?). From utterance to heart: The bias of language in speech and writing. Hanurdeducaflmaiketflew. fl. EST-HEB]. OLSON. EL. TOWER N.. e: I'ilLD‘I'r'iflD. A. (1935}. “my. ileum. dudlertming. New "I'tiltlt: Cambridge lilrriver.s'l.‘3.r Press. DYAME. S. {PETE}. The sensitive period for the acquisition of s nonnstrve phonological system. joumalql'Pgicba-J'flsguml'c Rermrcb, 5. 2151-235. PALMER. it. {19691. Henmemics- men-2mm new}: in Scatter-mariner, nosey. mama: amifiedamee Evansten. 11.: Northwestern Uriimrsity Press. PATKOWSKI. 1-1. (1994). The Critical age hypothesis in inter-[art- guage phonology In M. Yam {ELL}. First sndmortdhrtguage phonolomr (pp. 205—222]. San Diego. CA: Singular Publishing Group. PEARSON, PD. (1990). Foreword; Hole.I ] came to lutno'tliI about Beginning to read. In M. Adams, Beginning to read.- Tbettiing and learning aboutpn'utiw. Hn'ii}. Cambridge. MA: firmness. PEARSON. PD. (1997‘). The first-Grade Studies: A personal reflec- tion. finding Emmi: Qmmfy. 32.. 425—432. F'l-I.ll_II".‘.i:I S.U. (1983). fire inedfl'biie culture: Communicmon in classroom and community on the Warm Springs Midi-t Wind-1. New York: Longrnan. PRAISE AND CRE'T'JCISM FOR CJJNTON EDUCHTIDN PflLICY. (199?. August 1?}. The New their mates. See 1. p. 26. READEh‘CfiJ.E.. e Err-.mNE. DM. (1991'). Editorial: the First-Grade Sludies: The importance of literacy binary. Ming Research Quarterbl. 32:. 540—541. RJST. R. (197‘111. Student social class and teacher expectations. Harvard Eda-muons! Reuters; 40. 41145]. RUSH—HY. C. (1991. September 291. The man who took the State of Texas lo school. Hormone Chronicle Hmflagflinfil F. It}. ELITI'ER. M. {1933}. Sci-root effects on pupil progress: Research findings and policy implications. GeikiDet-eimmettt. 54. 1-29. SCI-IDOL FINANCE HUD NG: FWNTIFES HAIL LATEST WCTCIIR’Y IN SCHOOL CASE. (1993. September 26). HOW Chronicle. p. ALE. SCOVEL. T. [19331. .1 “were spite-lt- amber-ignich inquiry into the enricrripenpriferbumn w. New York: Harper & Row. SCHIBNE'R. 5.. IE COLE. M. (1931). maman Cambridge. MA: Harvard University Press. some. J. (19691. M were Cambridge. England: Cambridge University Press. STANGVICH. KI. {199131994}. Romance and reality. Item Teach 41". 230-291 THEE]... _T. (1991}. Multicultural literature and the polJIr'cs of rest:- tiort. Miners fibi'lege Record, 93, 411—643. TAHOE. Br. (1993). Winning It} rend and the spin doctors graci- ence: campaign to obsngeAme-rim’s mbrdabcmi' bowcb-i'l- int-rt fest-n to reed. Ul'bena. TL: Nadoml Cumeil enfTeeehers of English. TEALE. WH. {19615}. Home background Incl young children's liter- acy development. In Wit-1. Teale & l'-'_ Suizb‘y (Eon), Emergenrfflemcy EPP- 133-21316}. Nor-wood. NJ: Abieit. TEALE. W11. & SULZB’Y. E. Hm. Emerge-er lies-mtg». Nora-“God. NJ: .Ftblfl. 11W LEGISIAHVE BUDGET BOARD. (1994. December}. Trends Farrell: school accountability A report card on inflammation ty'ttre system. Austin, TX: Educational Enonomie Policy Ceruer. WOMPSQN. I. {1991}. Fweign aeeents retriei'led: The English pro- nunciation of Russia." irrin'ligflnts. Magma Lemming. 41'. [Tl—204. W. Mme. learning, andibegfiimn American child. Boston: Allyn E. Baron. WALEERG, l-IJ.I & T5311. 5. {1933}. Matthew elTects in education. American Edmfleneimrcbjmml. 25'. 559—513. WHITE, 5. (39%}. The recent work quurgen Halter-mas: Reason. justice. and modularity. Cambridge. England: Cambridge University PIES-G- WIIKINSCIN. 1.11.15. {199B}. Dealing with diversity: Achievement gaps in reading literacy among New Zealand students. Rpm-ting Mega: gum-reel}; 23. led—16?. 'WILIJNSK‘I'J. (1995). lean-ting to divide the twirli- Education at Empire: ext-cl. Minneapolis: University of l'r'linrlfiotz Frees. WILLIS, ILL, & MS, VJ. {1991]. Expuuiing the Mundan'es: A reaction to the First-Grade Studies. Mug Research Queue-i}; 32. 459-445. THE WINNERS & LDSERS £1? '97:": JOB SEEKERS AND SCHOOL KJDG GMNED. WHILE WEE“ PEIme TD-DK A l-IIT. (199?. December 3-1}. The Wallitrefljoermal, p. C1. W00. E. (1995. May 513'}. Smdents’ spelling gets an F. Californians are appalled by rentlts el ‘whole hnguage‘ mettle-d. mm Climrtr'ole; p. ME. YAVAS. M. (19913}. Phonology-I: Wee: and disorders. Sari Diego, CA: Emir: Publishmg Group. Received July 15. 1993 Revision reuecived November 2,. 1993 Accepted December 9. 1995 ...
View Full Document

This note was uploaded on 02/07/2010 for the course C&I CI550 taught by Professor Markdressmen during the Fall '07 term at University of Illinois at Urbana–Champaign.

Page1 / 28

Dressman_Research_Evidence - Mark Dressman Uniserst'flr...

This preview shows document pages 1 - 28. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online