Democratising the classroom: a literacy pedagogy for the new generation

South African secondary students see themselves as a ‘new generation’, the first to come of age in the democratic nation. They are intelligent, politically aware and highly motivated, but very few currently stand a chance of achieving their goals of further education and professional careers (Taylor, Muller, Vinjevold, 2003). If the new South Africa is to realise the possibility of a just society this situation urgently needs to change. But we are hamstrung, not just by the history of schooling in South Africa, but by classroom practices that have evolved in western education systems to reward the elite and marginalise the majority. This paper contends that the basis of inequality in the classroom, and hence in the society, is in students’ differing capacities to independently learn from reading, which is the fundamental mode of learning in secondary and tertiary education. Whether teaching practices are promoted as ‘learner-centred’ or ‘teacher-centred’ has little impact on the central problem of students’ differing capacities to engage in and benefit from them. This problem can be overcome if we focus squarely on teaching all learners in a class to read and write the texts expected of their level and area of study, as part of everyday teaching practice. I argue here that democratising the classroom is the primary condition for achieving the kinds of educational outcomes needed to build a democratic South Africa, and outline a literacy pedagogy that can enable us to do so. Learning to read: reading to learn The goal of the paper is to describe a methodology for teaching reading and writing that has been developed in a long term action research project with teachers in Australia at all levels of education, from early primary through secondary to tertiary study, across curriculum areas. The methodology, known as Learning to Read: Reading to Learn, has been developed in response to current urgent needs, particularly of Indigenous and other marginalised learners, to rapidly improve reading and writing for educational access and success. To this end it draws on three theoretical traditions: a Vygotskyan model of learning as social process, a Hallidayan model of language as text in social context, and a Bernsteinian model of education as pedagogic discourse. These theoretical foundations are integrated in a set of teaching strategies that 132 Journal of Education, No. 37, 2005 have been developed with teachers to be optimally practical in diverse classroom settings, and optimally practicable for teachers to acquire and use as part of their ordinary practice in their grade or curriculum area. The strategies have been independently evaluated as four times as effective as other literacy approaches at accelerating reading and writing development, capable of improving learners’ reading ability from junior primary to secondary levels within one year (McRae, Ainsworth, Cumming, Hughes, Mackay, Price, Rowland, Warhurst, Woods and Zbar, 2000; Carbines, Wyatt and Robb, 2005; Cullican, 2006). They are currently being applied in primary, secondary and tertiary contexts in Australia, Africa and Latin America, with learners from a wide spectrum of language, cultural and educational backgrounds. However before outlining the strategies I first need to address the broad educational context in which they have been developed and applied, and the theoretical bases from which they have evolved. Tools for democratising the classroom: Bernstein, Vygotsky and Halliday I will take as a theoretical starting point Bernstein’s model of teaching and learning as pedagogic discourse. Bernstein described pedagogic discourse as including two dimensions: “the discourse which creates specialised skills and their relationship to each other as instructional discourse, and the moral discourse which creates order, relations and identity [as] regulative discourse. . . the instructional discourse is embedded in the regulative discourse, and the regulative discourse is the dominant discourse” (1996, p.46). Crucially Bernstein saw these not as separate entities, but as aspects of a single process: “Often people in schools and in classrooms make a distinction between what they call the transmission of skills and the transmission of values. These are always kept apart as if there were a conspiracy to disguise the fact that there is only one discourse. In my opinion there is only one discourse, not two, because the secret voice of this discourse is to disguise the fact that there is only one” (ibid.). If we accept Bernstein’s view, one implication is that the dominant function of pedagogic discourse is not so much transmission of skills and knowledge, which is what we generally assume we are teaching, but rather of ‘order, relations and identity’. What then is the nature of this order, these relations and identities? I want to suggest that these are continually apparent to all teachers in all classrooms in every day of our practice. The dominant moral order in our classrooms is one of inequality. Teachers are confronted by this Rose: Democratising the classroom. . . 133 inequality from the day we first walk into a classroom, ill-prepared by our training to manage it, let alone overcome it. Every one of the teachers we work with in our in service programs, from early years on, report that a minority of learners are consistently able to actively engage in classroom activities, to respond successfully to teacher questions (the primary means by which we interact with our students), and to succeed in assessment tasks. Another group are sometimes able to actively engage, to respond to questions, and achieve average success, while a third group are often unable to engage, rarely respond, and are frequently unsuccessful in tasks. Relations in other words, between learners within every classroom and school, are unequal. As a result the learner identities that are produced and maintained by the moral order of the classroom and school are stratified as successful, average or unsuccessful. This inequality is universally construed at all levels of education, whether overtly or not, as differences in learning ‘ability’. The entire educational edifice of assessment, progression and specialisation is predicated on this assumption. The naturalisation of inequality as differences in ‘ability’ serves to internalise these identities, so that successful learners come to experience schooling as their pathway to the future, while unsuccessful learners eventually come to experience it as irrelevant, even alienating. Education and socioeconomic inequality I now want to propose that this moral order within the classroom, these unequal relations between students, that specialise learner identities as successful, average or unsuccessful, is the primary engine in modern industrial societies for reproducing socioeconomic inequality. In other words the evolved (rather than designed) function of pedagogic discourse in modern education systems is to reproduce an unequal social order. The broad function of instructional discourses embedded in this regulative function is then to specialise occupational roles as ‘professional’, ‘vocational’ or ‘manual’ (Rose, 1998). That is the instructional function of pedagogic discourse is to specialise economic roles, while its regulative function is to naturalise this specialisation by formation of differing learner identities. This relation drawn between classroom inequality and social hierarchy is materially supported by statistics of educational outcomes. Over the past twenty years there are has been relatively little change in outcomes in Australia, as displayed in Figure 1. The high proportion of Australians with no post-school qualifications has decreased marginally from over 60% to above 50%, while the proportion of those with bachelor or higher degrees has 134 Journal of Education, No. 37, 2005 increased from about 7% to 17%. In between, the proportion with vocational diplomas or lower certificates has remained constant at about 30% (ABS, 1994; 2004). Given the resources poured into the Australian education system over this time, and the energy devoted to educational debates and changing teacher practices, a 10% change in 20 years represents a very slight improvement. Yet even this change is much less than it appears, as a large fraction of increased post-school qualifications are basic certificates and on the job-training, while much of the increase in bachelor degrees represents amalgamation of technical colleges with universities, and reaccredidation of diplomas as degrees. Figure 1: Educational outcomes in Australia 1984-2004 The constancy of these proportions is significant from two perspectives. One is that they reflect occupational strata in developed economies, with a relatively small professional elite, a larger segment of vocationally trained trades people, and a larger pool of on-the-job trained or unskilled manual workers. The latter unskilled occupational fraction has shrunk drastically in post-Fordist economies (Harvey, 1989), leading to crises such as youth unemployment. The failure of educational outcomes to keep pace with this socioeconomic change has focused attention on literacy in schools, and led to imposition by the state of testing regimes in an effort to force improvements. Secondly they strikingly reflect the proportions of groupings that teachers report in their classrooms, of successful, average and unsuccessful learners. These proportions vary from school to school and region to region, but they Rose: Democratising the classroom. . . 135 Concern over the outcomes of progressivist pedagogy has been echoed by diverse secondary and 1 tertiary educators I have met around South Africa, who are finding literacy levels deteriorating at a time when they need to be rapidly improving. It is now crucial that we look beyond the seductive rhetoric of progressivism to the reality of its outcomes for the most vulnerable learners. A sceptical eye needs to be cast over the cluster of associated practices and philosophies that oppose explicit teaching of school knowledge and school language in favour of self-discovery, including notions such as ‘language experience’, ‘whole language’, ‘process writing’, ‘discovery learning’, ‘reading circles’, ‘peer scaffolding’, ‘social constructivism’, and so on. Proponents of these ideas are often consummate persuaders but their primary interest is not in providing equal outcomes for all (cf. Muller, 2000). persist despite apparent major changes in prevailing pedagogic practices. The twenty year span of Figure 1 covers a period in which teacher training in Australia has been entirely dominated by the set of practices and philosophies known as progressivism. As progressivist practices are legitimated on principles of access and equity, of inclusiveness and ‘learner-centredness’, it is ironic that they appear to have had negligible effects on inequality of outcomes for the majority of less advantaged learners in this wealthy nation. It may be argued that at least outcomes have not gotten worse (as some commentators claim), but for those in the least successful group, our failure to improve outcomes has been an accelerating calamity. This includes those school leavers who would formerly have gone into unskilled manual labour, but today make up the 30–40% of unemployed young people in Australia (and many more in South Africa). Nowhere is this failure more calamitous than in Indigenous Australian communities, who were formerly excluded from educational access by racist policies, and desperately needed rapid improvement in outcomes to gain employment, manage their communities and negotiate with the colonising society. Instead our failure to educate recent generations of Indigenous students has resulted in unemployment rates of 60–90% in many communities, endemic intergenerational welfare dependency, and concomitant social disasters (Pearson, 2002; Rose, 1999). The experience of Indigenous Australian learners with the hegemony of progressivist pedagogy sounds an ominous warning for the many South African educators who have embraced its promise of liberation from authoritarian past practices. From the perspective of actual outcomes this promise now sounds increasingly hollow. 136 Journal of Education, No. 37, 2005 Bernstein refers to progressive and traditional pedagogies as ‘competence’ and 2 ‘performance’ models respectively (1990, 1996, Muller, 2000; Rose, 1999). The hidden curriculum of reading development I would like to suggest here that the ideological struggle between progressivist and traditional pedagogies is marginal to the core function of schooling to 2 service the needs of a stratified socioeconomic order, by reproducing occupational specialisation as professional, vocational and manual labour. The engine of this reproduction is not primarily in the overt content of the curriculum, nor in an emphasis on learner-centred or teacher-centred philosophies, but in persistent evolved classroom practices that engage and enable different learners unequally. The term ‘hidden curriculum’ has been used to refer to positioning of learners through ideologically loaded curriculum (e.g. Muller, 2000), or to ‘invisible’ forms of control characteristic of progressivist pedagogies (Bernstein, 1996). Here I would like to use the term in a very specific sense to refer to practices that construct, maintain and evaluate inequalities between learners. The content of this hidden curriculum is inequality in students’ abilities to participate and perform successfully. The process by which this is achieved is ordinary classroom discourse, including the ‘triadic dialogue’ of question-response-feedback described by many analysts as endemic to classroom interaction (see further discussion below). The superficiality of the progressive/traditional ideological conflict, supervening on the underlying iceberg of unequal ‘abilities’, is represented in Figure 2. Rose: Democratising the classroom. . . 137 Figure 2: Superficiality of progressive/traditional pedagogic conflict progressive traditional pedagogy pedagogy Hidden curriculum content: unequal ‘abilities’ process: classroom discourse From where does the inequality in ‘ability’ arise? Few of us now accept that its basis is in biology, in some as yet undiscovered differences in the structures of learners’ brains. Rather it is generally accepted as cultural in origin, and specifically in different kinds of preparation and support for the demands of school learning, provided by children’s primary socialisation in the home. The most obvious and relevant difference in this respect is in the experience of parent-child reading, of which children in literate middle class families 138 Journal of Education, No. 37, 2005 experience an average of 1000 hours before starting school (Bergin, 2001), whereas those from oral cultural backgrounds may experience little or none. I have suggested that parent-child reading before school is the first stage in a curriculum of reading skills that underlies the content and processes of the overt curriculum in each stage of schooling (Rose, 2004a). Children with this wealth of experience are in a position to benefit most from the next stage of the underlying curriculum – the literacy practices of junior primary teaching (whether these are construed as traditional or progressive), and rapidly learn to become independent readers. It is crucial that these children are independently reading with understanding and engagement by the end of Year 2 or 3, in order to be ready for the next stage of the curriculum in middle to upper primary, in which they learn to learn from reading, and to demonstrate what they have learnt in written assessment tasks. These skills are in turn essential for these learners to be ready for secondary school, in which the fundamental pedagogic mode is through independently learning from reading. Skills in learning from reading are rarely taught explicitly in upper primary or secondary school; rather successful learners acquire them tacitly over years of practising reading and writing the overt curriculum content in class and homework. The accelerating volume of this content in the secondary years forces successful students to develop the skills they will need in tertiary study for independently reading academic texts, and reproducing and interpreting what they have read in assignments. So each stage of the reading development curriculum, from parent-child reading onwards, prepares learners with the skills they will need for the next stage. But as these skills are not explicitly taught in the following stage, what learners are evaluated on are actually skills they have acquired in the preceding stage. That is junior primary teaching evaluates children on reading orientations they have acquired in the home, upper primary practices evaluate them on independent reading skills acquired in junior primary, and so on. Those learners who have acquired skills in each preceding stage are continually affirmed as ‘able’ in the next stage, while those learners who have not acquired the skills are evaluated as ‘unable’. Evaluation is not simply or primarily through formal assessments, but continues relentlessly in the form of ordinary classroom interaction, in which teachers ask questions of the class that serve to differentiate learners on their ability to respond successfully. By these means, relations between learners are constructed as unequal from the very beginning of schooling, and their identities are continually reinforced as successful, average or unsuccessful. This reading development curriculum Rose: Democratising the classroom. . . 139 underlying the overt content focused curriculum sequence of schooling is diagrammed here in Figure 3. Figure 3: Reading development sequence It must be emphasised that the reading development curriculum has evolved as mass schooling has emerged in stages from earlier systems for preparing elites for professional training; it is not a designed system, there is no conspiracy, and the overwhelming majority of teachers would prefer that all their students were successful. But as it underlies the overt content curriculum, beneath the notice of practitioners, and is acquired tacitly by elite learners, it serves a double function. One is to prepare the successful few for university, the other is ensure that other learners do not acquire the same skills. The first function is achieved by forcing able learners to continually practise reading and writing across the accelerating content curriculum; the second is achieved by not allowing time to teach these reading and writing skills explicitly. However as with Bernstein’s pedagogic discourse, this evolved double function is really one, to reproduce a stratified social order. The good news is that it is possible for all learners to rapidly acquire skills in reading and writing at any stage of the curriculum, by teaching them explicitly instead of leaving them for tacit acquisition. It takes successful learners six years each of primary and secondary schooling to acquire these skills, precisely because they are not taught explicitly. But we have demonstrated that they can be acquired by the weakest of secondary students in a year of explicit teaching, with a mere 2 or 3 140 Journal of Education, No. 37, 2005 The origin of the incremental learning model may be associated with the medieval approach 3 to teaching classical languages, beginning with the smallest units of syntax and building up in prescribed steps. lessons per week. The Reading to Learn strategies are designed to by applied at any point in the reading development sequence, as either repair or part of ordinary teaching practice. The Vygotskyan model of social learning Although progressivist philosophies are rhetorically opposed to so-called traditional practices – as learner-centred vs teacher-centred, inclusive vs exclusive, wholistic vs atomistic, discovery vs rote, and so on – they have failed to significantly change outcomes because they share fundamentally the same model of learning that is tacitly assumed in schooling in general. In this model, learning is presumed to happen within individuals in increments as they master one step after another. The incremental learning model did not originate as a theory but as a tacitly held view that appears to have evolved with the vocation of school teaching. That is its context of evolution was a 3 pedagogic discourse that produces unequal order, relations and identities. It can be contrasted with tacit models of learning that underlie other pedagogic domains. Examples include trades training, where apprentices are expected to emerge with a common set of vocational skills and identities, and are explicitly supported to achieve these outcomes, from the ‘outside in’, by repeated modelling and practice (Gamble, 2003); or in the home where children are expected to acquire a common set of linguistic and cultural skills and identities, and are explicitly supported by their parents to acquire them through repeated modelling and practice (Painter, 1984; 1996; 1998; 2004). As teaching became professionalised the incremental learning model was theoretically legitimated and formalised in Piagetian child psychology which privileges relations between learning and innate developmental stages, i.e. that learning takes place from the ‘inside out’ (Piaget, 1928). On the other hand the Piagetian model of innate development did lead to rejection of evolved traditional teaching practices that were construed as teacher dominated, i.e. from the ‘outside in’. Instead learners are given tasks matching their assessed ‘ability’ level, and learning is assumed to occur as individuals do these activities. This model demands that learners are continually evaluated to assess their readiness for advancement. In traditional practices these assessments may inform and legitimate streaming into different ‘ability’ classes. In progressivist practices they may inform individuated learning activities that differentiate Rose: Democratising the classroom. . . 141 learners within classes, and this hierarchical differentiation is then legitimated as ‘learner-centred’. The incremental learning model is schematised in Figure 4, in which learners complete a series of independent tasks at their various ‘ability’ levels, leading to summative assessment tasks. As the rate of development of less successful learners cannot exceed the rate of more successful learners, this model ensures that the ‘ability’ gap can never be closed. Figure 4: Incremental learning model: unequal outcomes In such an individuated view of learning, teaching is wholly constrained by the independent competence of the learner. If some learners fail where others succeed, there is little teachers can do beyond individual ‘remediation’. Since learning is assumed to occur through independent activity, and assessment is continuous, learning activities and assessment tasks are not clearly differentiated. What teachers perceive as learning activities, learners may perceive as evaluation tasks, particularly those learners who are least successful. Thus all manner of activities, from formal reading and writing tasks, through maths and other short exercises, to the question-responsefeedback routines of classroom interaction, all serve to produce and maintain learner identities as more or less successful, no matter what their instructional intent. 142 Journal of Education, No. 37, 2005 An entirely different view of learning is the Vygotskyan model (Vygotsky, 1978; 1981), which claims that learning takes place in the ‘zone’ between what learners can do independently and what they can do with the support of a teacher. We can apply this view to reflect on traditional and progressive pedagogies which depend on learners’ independent competence. In teachercentred modes this can take the form of presenting information to learners and relying on them to assimilate and use it independently. In learner-centred modes a context is provided in which learners are expected to ‘discover’ concepts for themselves. But in the Vygotskyan view learning takes place in both modes only insofar as learners are supported by a teacher or by a text that mediates the teacher’s support. Teacher-centred activities clearly provide sufficient support for some learners to acquire the information presented, but insufficient support for others. Learner-centred activities provide a modicum of support for all learners, but the level of task is higher for some and lower for others. Both sets of practices advantage more advanced learners, as they are pitched just beyond their independent competence. Both also disadvantage less advanced learners: teacher-centred ones because they are too far beyond their independent competence and provide insufficient support to bridge the gap; learner-centred ones because they ‘dumb down’ tasks below the levels achieved by more successful learners, so that their rate of incremental development falls further and further behind that of the more successful learners. In contrast to both these sets of practices, the Vygotskyan model suggests that a teacher can potentially support learners to operate at a high level no matter what their independent ability. The Learning to Read: Reading to Learn pedagogy assumes this possibility, but takes it further to support all learners in a class to simultaneously operate at the same high level. In this model the teacher is neither simply an authority presenting information, nor simply a facilitator managing a learning context, but a guide providing what Bruner has called ‘scaffolding’ (first in Woods, Bruner and Ross, 1976; cf. Mercer, 2000; Wells, 1999). In the Reading to Learn methodology, scaffolding supports all learners to do the same high level tasks, but provides the greatest support for the weakest learners. Rather than developing in incremental steps, learners acquire independent competence through repeated practice with high level tasks, and the scaffolding support is gradually withdrawn as learners take control. This then is the principle by which an unequal moral order can be transformed into a democratic classroom, where successful learner identities can be distributed equally to all students. Rose: Democratising the classroom. . . 143 The Hallidayan language model and reading The goal of democratising the classroom is not a utopian dream. It is basic practice in the Reading to Learn program, made possible by the contribution of Halliday’s functional model of language to understanding and so explicitly teaching the tasks of reading and writing across curricula (Halliday, 1975; 1978; 1993). Central to Halliday’s theory is the notion of realisation, where meaning is realised as wording (i.e. ‘expressed/ symbolised/ manifested’), and wording is realised as sounding or lettering. Theories of reading in early schooling tend to be polarised between those that focus on comprehension of meaning, often advocating ‘immersion’ of learners in whole texts (‘whole language’), versus those that advocate explicit teaching of sound-letter correspondences, followed by words, phrases and sentences (‘phonics’ and ‘basal readers’). In Halliday’s stratified model of language, this polarisation dissolves into different perspectives on the same phenomenon, from the stratum ‘above’ of meaning or discourse semantics, and from the stratum ‘below’ of sounding and lettering or phonology/graphology (Halliday 1996). It is the stratum between, of wording or lexicogrammar, that is typically conceived as what we are reading, since the written page consists of words organised into sentences. The acrimony in reading theory is over whether it is primarily ‘decoding’ sequences of letters, or ‘predicting’ sequences of meanings, that enables us to read words. The answer flowing from the Hallidayan functional model is of course both. Layers of structure in these three language strata are represented schematically in Figure 5. Figure 5: Complexity of the reading task by strata and rank 144 Journal of Education, No. 37, 2005 The medium of expression, of sounding versus lettering, is an obvious difference between speaking and writing, so explicit teaching of reading has traditionally started with teaching the graphic medium. But Halliday (1989) has also shown us significant grammatical differences between spoken and written modes of meaning, between the ‘recursive’ structures typical of speech and ‘crystalline’ structures typical of written sentences. Essential for recognising these differences is his model of grammatical ranks: while a written sentence may appear visually as a string of words, its meanings are also organised in intermediate ranks of word groups or phrases. For example the sentence A frog was swimming in a pond after a rainstorm consists of four word groups, denoting who its about, what they were doing, where and when. Where lexical ‘content words’ tend to be sparsely strung out in speech, in writing they are densely packed into groups within each sentence, as well as into technical and abstract words. The practice of packing complex meanings into abstract wordings is known as grammatical metaphor (Halliday, 1994; Martin and Rose, 2003; Rose, 2000b). Where experienced readers are able to automatically process such lexical density, inexperienced readers may labour to ‘unpack’ dense wordings, often without success. Likewise, a word appears visually as a string of letters, but these are actually organised in intermediate ranks of syllables and their components. A layer of structure above the letter is acknowledged in phonics approaches to reading, as letter ‘blends’ that are drilled in lists of sound-letter correspondences. But the sounds associated with letter patterns in English vary with the particular word in which they occur (the ‘ough’ pattern is one obvious example), and with their structural position in the syllable, as onset (e.g. ‘thr-’) or rhyme (e.g. ‘-ough’). The entire English spelling system is thus very complex, but like all language systems consists of regular predictable contrasts (Mountford, 1998). These can be learnt, not from drilling sound-letter paradigms and sounding out words, but only from recognising recurrent instances in meaningful discourse, as we learn language in general. Experienced readers recognise words by visually processing letter patterns, whereas weak readers often struggle to sound out words letter-by-letter, a strategy encouraged by phonics approaches. But it is not through processing letter patterns alone that we recognise written words; while the spelling system is complex, the systems of meaning that wordings realise are immeasurably more so, and it is equally our experience of these systems that enables us to read. Again there are intermediate layers of structure in the discourse semantic stratum, between the sentence and the text, in particular the stages that different genres go through to achieve their purposes, as well as shorter phases of meaning within each stage that are more Rose: Democratising the classroom. . . 145 This is a richer interpretation of reading than can be offered by cognitive reading theories, 4 because it is grounded in large scale detailed analyses of how language actually makes meaning, rather than hypotheses of how the mind/brain works. In contrast cognitive theories are necessarily speculative, as ‘mind’ can only be observed indirectly through behaviour. Cognitive theories may converge with the model presented here insofar as they advocate strategies based on observation of successful reading behaviour, and because they are working with the same phenomenon – language. The major difference is that they skirt around this phenomenon, touching on it at certain points, but at the heart of their discourse there lies a vacuum left by the absence of an articulated theory of language in social context. The same absence lies in the discourse of ‘New Literacies’ theorists, whose perspectives tend to the social rather than cognitive (e.g. Gee 1998; Street 1999), but who lack a detailed understanding of the object of their expertise – language – and how to