ROW02074

Patricia M. Rowell, University of Alberta

Margaretha Ebbers, Edmonton Public Schools

Abstract

Many researchers are probing school science from a sociocultural perspective, giving particular attention to the ways in which language events of the classroom shape the school subject, as well as how the school subject, in turn, shapes the classroom discourse. The recognition by researchers of literacy practices in science communities is not, for the most part, shared by either teachers or policy makers in the Canadian province of Alberta. While one of the features characterizing the work of scientists is the construction of tentative explanations and the social validation of these by a science community, scant attention has been paid to inducting students into such discursive practices. Although the mandated provincial curriculum is framed by science inquiry, limited support is provided for the literacy practices needed to move from description of events and phenomena to the construction of explanatory models. In this paper, we draw on case studies of school science in two Grade 6 classrooms. Data are derived from field notes made during observations of lessons comprising two instructional units (34 lessons), transcripts of audiotaped lessons and teacher interviews, and copies of student work. We focus analysis on classroom interactions contributing to construction of explanations.

More than a decade ago, White and Welford (1988) noted that "learning how to use language (italics in original) well in all areas of the curriculum becomes, at best, a matter of osmosis; at worst, something that fails to happen" (p.34). The authors were commenting on a small selection of student writing in science, when students were asked to make and interpret observations. And so, it is of concern to read a similar call today that "No matter what the subject, the people who read it, write it, and talk it are the ones who learn the best" (NCTE, 2002). And of the references cited in this call, the most recent are dated 1990. One might reasonably ask what has happened since these earlier admonitions to heed the role of language in learning science in school? Research in science education in the past decade has focussed to a large extent on the concepts and processes of science inquiry, and ways of helping students, as individuals, acquire the conceptual tools needed to construct explanations for phenomena and events in the natural world. What has been largely neglected is the way in which language mediates the participation of students in the classroom community when oriented to learning science.

As researchers, we approach the teaching and learning of science inquiry practices by acknowledging the role of language in learning communities. The use of language is much more than an exchange of information; language constitutes the discourse which mediates the social practices of specific communities (Driver et al., 1994; Lemke, 1995; O'Loughlin, 1992: Swales, 1990; Wells, 1995). As such, discourse is language-in-use (Gee, 1996, 1999); it is socially situated in sustained social relations between participants in face-to-face interactions or between authors and readers in written texts (Hicks, 1996). From this viewpoint, ways of doing things in particular communities (cultural) and ways of communicating (social) pervade specific situations. Literacy practices are "primarily courses of social action" (Luke & Freebody, 1997).

In this article, we first report on case studies in the science lessons of two Grade 6 classrooms before considering the implications of the findings for the role of literacy in school science.

Science may be thought of as cultural activity in which participants strive to relate theory and data to support explanatory models. In an explanatory model, relationships are articulated to provide a possible explanation of how a phenomenon occurs. The relationships are constructed by people in a deliberate fashion to account for the features of the phenomenon. Such models are not derived from empirical observations, but empirical observations may be used to generate evidence supporting an explanatory model. Explanatory models may then be used to predict behavior of a phenomenon under specified conditions, giving rise to test situations (Driver, Leach, Millar & Scott, 1996). Thus in science communities, members engage in literacy practices oriented to the construction and articulation of explanations which move beyond mere description or correlation of observable variables; the culture of science values the generation of theories.

Science may also be regarded as a social activity, in that theories (explanatory models) and data are shared among communities of scientists, and validated in light of acceptable evidence (Norris & Phillips, 1994). Members of science communities share in socially ways of using language, of thinking, and of acting; that is, they participate in a shared discourse (Gee, 1996). The discursive practices of these communities entail using language to transform data into evidence, and then to argue persuasively the relationships between evidence and an explanation. Talk, reading and writing contribute to the practices of argumentation in which a cultural discourse recognized as science is transmitted, transformed and contested (Luke & Freebody, 1997).

Policy statements (AAAS, 1993; National Research Council, 1996; Council of Ministers of Education Canada, 1997) about school science articulate the need for students to be engaged in practices which model those of science communities. In curriculum documents, teachers are advised to encourage students to formulate questions about natural phenomena and to engage in investigations which will offer data in pursuit of explanations. Students are to be taught how to interpret data to generate evidence which, in turn, is used to support or refute explanations.

Although Munby (1973) and Barnes (1976) drew attention to the critical role of language in communicating and structuring learning in science, it is relatively recently that researchers have begun to explore how the discursive tools of oral text (Gallas, 1995; Lemke, 1990; Michaels & Sohmer, 2000), print text (Ford, Palincsar & Magnusson, 2000; Heselden & Staples, 2002; Norris & Phillips, 1994; Shymansky, Yore & Good, 1991; Sutton, 1989; Wellington, 2001) and written texts (Christie, 1996; Derewianka & Schmich, 1991; Haneda & Wells, 2000; Keys, 1994, 1999, 1999a; Staples & Heselden, 2001) contribute to the practices of school science.

Many elementary science programs adopt an inductive relationship between observations and explanations, assuming that explanations may be structured directly from observational data in the form of empirical generalizations (Driver et al, 1996). However, some elementary science programs include topics, such as flight, which do not lend themselves to the gathering of data from which explanations can be empirically generated. Such topics require the use of constructs (such as lift, drag) and relationships between them (concepts) to be imposed on phenomena in order to make sense of them. Theoretical entities (such as molecules) may be posited to account for the features of a phenomenon in the form of an explanatory model. Hence, explanatory models are not derived from empirical observations; they are conjectural. Support for such explanatory models, however, is dependent on the generation of evidence through inquiry and subsequent persuasive argumentation. These may include the use of a model to predict the behavior of a phenomenon under specified conditions, giving rise to test situations (Driver et al., 1996).

In school science, then, how do literacy practices in the instructional discourse scaffold students in appropriating scientific discourse? There is limited research reported on the discourse and practices of teachers as they formulate explanations in the science classroom (Ogborn et al., 1996; Russell & Munby, 1989; Southerland et al., 2001), on the explanations generated by students (Driver et al., 1996; Dagher & Cossman, 1992) and explanations in print resources (Unsworth, 2001). Much of this work has been carried out in secondary schools rather than elementary. In this paper, we report on case studies in two grade 6 classrooms where a focus of instruction was the adaptations of birds for flight. Our examination of the instructional discourse is guided by the questions "What kinds of explanations of bird adaptations for flight were constructed by students?", " How was the construction of explanations modelled by the teachers?", and "How did the instructional discourse scaffold construction of these explanations?

This qualitative study was carried out in two classrooms in a western Canadian urban school district during the teaching of two instructional units: Air and Aerodynamics and Flight, in two successive school years. A Program of Studies (Alberta Education, 1996) for elementary science is mandated by the provincial government. Five instructional units are prescribed for each year in Grades 1 - 6, of which four are designated as science inquiry. Specific learner expectations are set out for each instructional unit in this program. Examples of such expectations in the Air and Aerodynamics unit are that students will 'identify adaptations that enable birds and insects to fly', and 'describe the means of propulsion for flying animals and for aircraft'.

The government does not mandate instructional resources, and there are no prescribed textbooks. Instead, a collection of teacher and student print resources are identified as 'authorized' for use in classrooms, their selection being at the discretion of individual teachers, in consultation with colleagues and school administration. Following observations of the Air and Aerodynamics and Flight instructional units in the first year of this study, we consulted with the teachers and prepared modular resources (including collections of children's books about birds and flight) intended to support language-oriented activities which complemented the manipulative activities developed by the local school district. The teachers were invited to make use of these resources within the umbrella of the language education program.

In an ethnographic mode, data collection entailed audio-taping each lesson in each classroom in the units (more than 30), supplemented by observer field notes. The length of the lessons ranged from 60 - 90 minutes. Interviews with the teachers at the beginning and end of the study were also audio-taped and transcribed. Notes were made of informal conversations with teachers during the instructional period. Interviews with the students were conducted at the end of the unit and their written work was copied. This report draws on the transcripts of tapes and field notes from the lessons which featured the adaptations of birds with respect to flight, from interview data, and from written work samples that correspond with these lessons.

Through an interpretative analysis of field notes, transcripts of interviews, student written work and transcripts of lesson interactions, we have explored the ways in which explanations regarding the adaptations of birds with respect to flight were generated. In this analysis, the focus is directed to the pedagogical events constructed by the teachers and students and on the engagement of the students in a variety of language and inquiry practices geared towards collecting evidence that could be used to help frame an explanation of flight by birds. Transcripts of classroom interactions have been coded and interpreted for the accomplishments of the interactions, and for the demands of literacy in science and science inquiry in attaining these accomplishments.

The two teachers worked in the same school, located in a middle class subdivision of the city. The more experienced teacher, Hannah, had taught for 8 years (4 with grade 6) and considered language education to be her area of expertise. Hannah was an active member of the local chapter of the International Reading Association, and had served as a board member in various capacities. The students in Hannah's Academic Challenge class (12 girls, 17 boys) had all been identified as performing at an advanced academic level and had met specific district entrance requirements for such a grouping. Lara, with 4 1/2 years' school experience, taught music to several grades in the school. There were 27 students (9 girls, 18 boys) in Lara's Grade 6 class. Neither teacher regarded herself as a science specialist.

The lessons dealing with bird flight were embedded within more extensive units entitled Air and Aerodynamics and Flight . As an introduction, the students read a children's book, The Puzzle of the Dinosaur-Bird (Schlein, 1996), which recounts the varied arguments presented by different scientists to account for the fossilized remains of the Archeopteryx. In Module 1 (Winging It), students and teachers investigated bird adaptations for flight through a collection of data from primary (observations of feathers, birds) and secondary (nonfiction texts) sources. In this report, we focus on strategies used by the two teachers to encourage the construction of students' explanations of three adaptations of birds for flight (see Note 1).

At the conclusion of the instructional unit, we asked students during individual interviews to explain why birds could fly and they could not. The response of one student was:

As this student suggested, there are many features which contribute to a bird's ability to fly. We have chosen to focus on the three adaptations mentioned most frequently by the students during these interviews:

Birds have hollow bones

We sorted the students' explanations into three categories on the basis of how the students attempted to link a particular adaptive feature to its usefulness in either decreasing or increasing a particular force in flight. We have termed these as descriptive explanations, relational explanations and explanatory models; examples of students' explanations in each of these categories are provided in Table 1.

Descriptive explanations: By far the most prevalent was the use of description as explanation. Students identified an adaptation but did not attempt to articulate its relationship to flight.

Relational explanations: In this category, students identified a particular adaptation and made an attempt to articulate its function with respect to flight. At times, these links were oriented towards cause and effect. This cause and effect was either clearly described such as "They have hollow bones which are lightweight so it is easier to get airborne" or implied, as in "Their hollow bones enable them to be light". At other times, the link was made simply by adding the function without articulating a cause and effect relationship as in "Feathers make it streamlined".

Explanatory Models: Students seldom articulated the relationships between the phenomenon of flight and the forces acting on the bird. Only on rare occasions did students generate an explanatory model. These explanatory models were more often constructed orally during class discussion, rather than in written responses. In these explanations, the students linked an adaptive feature and its accomplishment with respect to flight directly to force being generated or opposed.

In both classes, attributes of feathers, bones and chest muscles were emphasized in oral interactions. Lara notes the light weight of a bird's skeleton:

L: The skeleton is very lightweight.

If you don't have that it's very lightweight, just write that in. (April 4)

Descriptions of feathers included labelling different types of feathers, as well as labelling features of individual feathers. Early in the instructional unit, Lara introduced a description of chest muscles:

They're flying. Powerful. (Feb 23)

Students also contributed to the compilation of these descriptions, as in the following example from a student in Hannah's class:

S: So if you think about it, the endurance they [hummingbirds] would have to have in their arm muscles, they would be pretty strong compared to, say, an elephant.

(March 7)

This student went beyond description of bird muscles in correlating their strength to the endurance needed for rapid flapping of the wings. Lara also pointed out to students that having a body that was not heavy would be helpful for flight:

L: Fewer bones than reptiles or humans. I think that makes sense to me.

Even though the bones have hollow spaces and they're very strong and they have no teeth, they have all these things, the bones still have some weight, so the less you have the better to fly? (April 4)

And Lara also correlated streamlining with the notion of being aerodynamic:

L: The reason that the feathers are so connected is because we need to streamline. The bird needs to make itself more aerodynamic. (April 4)

Even more specifically, Lara moved to link adaptive features with the forces of flight, such as drag or thrust:

L: How does the bird get its thrust? Ben?

B: They can run on water.

L: It might be able to get a running start.

M: They flap them.

L: Yes, they flap their wings together, and that creates the thrust for the bird to go forward, so that's a really important part of wings, right?

They create thrust. (April 4)

In Hannah's class, students also offered connections to the forces of flight, although Hannah seldom picked up these conections:

S: [primary feathers] propel the bird forward. (March 7)

S: [There is] less resistance; they reduce drag. (March 7)

One of the student tasks in these lessons was to gather information from a collection of children's nonfiction books about the role of birds' adaptive features in flight. Lara introduced this task as a search for information "about birds in general; their bones, their lungs, their chests and their wings". She did not emphasise the role played by these features in achieving flight. After working on this task, Lara asked students to report on "what they had found out", inviting students to "tell us about feathers". And so, while Lara was clearly hoping to move students from descriptive to relational explanations and explanatory models, she did not encourage students to examine these explanatory forms offered in the books.

Students used the book collection as a source of data in their investigations and often read sections to one another when relevant information was identified. The students were encouraged to use more than one source when locating information that could be used as evidence. The passages of text could have been used as models for converting information into an explanation or argument. The following passage (Brinkley, 2000, p. 14) was read aloud by Hannah during discussion of different types of feathers. Note the point at which Hannah stopped reading:

Hannah stopped reading here; but the next sentence reads......

Hannah stopped reading aloud precisely at the point in which the text linked the features of feathers in the bird's wing with the force of drag. And, in so doing, constrained the explanation to a relational one rather than moving on to an explanatory model.

While passages were read regularly by the students as they collected information, and on occasion by the teacher in order to facilitate communal gathering of information, there was no time in which the students examined specific features of the text structure in order guide their growing ability to create more complex explanations. In addition to the text sets of children's literature, the teachers distributed to students notes prepared as backgound information for teachers (Edmonton Public Schools, 1996). These notes consisted largely of a series of descriptions of bird adaptations for flight (wings, feathers, skeleton, muscles and breathing system), but with a few models of relational explanations or explanatory models. For example:

In any collection of print resources, it is anticipated that differing views and interpretations may be found. Text sets thus afford opportunities to debate competing ideas, their sources and validation. Prior to using the text set, Hannah elicited students' ideas about reasons for using multiple textual resources. Students suggested that "not every source is the same", " [information is] not always right", "Some might have more information than another one", "Some books are older", "Not everyone knows everything". Hannah did not elaborate on these suggestions nor suggest any strategy for working with multiple sources. While students in this class were willing to acknowledge text resources as sites for selection and validation of data or evidence, their teacher was hesitant to do so.

In Lara's class, it was apparent that students had encountered strategies for dealing with conflicting information.

L: What if you find different information in two different sources?

L: Did you hear that, Victoria?

A: If you find sources that are not the same, you go with the majority.

L: You go with the majority.

Usually a more recent book will have the most recent ideas. (April 4)

The process of evaluating information in nonfiction trade books has the potential for developing a critical stance in reading. Rather than assessing the blind authority of texts based on 'a majority' vote, discrepant information demands inquiry into competing ideas and the evidence supporting them. Multiple print resources afford opportunities to examine data and ways in which the data have been generated and transformed. However, for the teachers in this study, achieving some degree of certainty with respect to the 'correctness' of the information was a higher priority than probing the sources of discrepancy.

Lara made sure that important phrases were part of the students' written explanations by coaching for their inclusion.

L: So that's one part that a bird has that we will never have, that will help it to fly.

Prior to working on written tasks, Lara used whole class question and answer exchanges to foreshadow required written responses. Once the students had completed their written tasks, Lara again emphasized the key explanations required. This is demonstrated in her following remarks, made in her review of the task:

L: And the bird goes up in the air.

Lara repeatedly checked to make sure that the students were recording what she considered to be important information. For example, she prefaced the modeling of an oral explanation regarding the usefulness of streamlining by asking students to record a definition of streamlining. This definition was written on the board and the students were directed to copy it into their notebooks.

Lara encouraged the students to expand their descriptive phrases and move to relational explanations. Sometimes she tagged on the function of the adaptive feature mentioned by the students, as in the following excerpt:

L: Contour feathers are helping to streamline. Can you see that?

Other times, Lara prompted students to link the adaptive feature to flight forces:

S: Drag. (April 3)

Lara also reminded the students not to be satisfied with just a descriptive explanation. When a student suggested that a bird has strong chest muscles to be a strong flyer, Lara responded with the following questions:

L: What are those chest muscles moving? In order to flap the wings, right?

Consistently, Lara incorporated students' ideas into the discourse, linking suggestions to the accomplishment of flight, and articulating connections. For example, when a student suggested that birds flap their wings too fast to see the movement, Lara responded:

And, whenever a student attempted to construct an explanatory model, Lara provided support:

Because that is what a bird is constantly working against. (April 3)

Not only did Lara scaffold individual students so that they had "a good way of thinking about it", she reminded the class repeatedly that birds are always working with flight forces.

Hannah did not scaffold construction of explanations needed for the written tasks, as in Lara's class. However, there were several occasions when oral explanations were both modeled and communally constructed. Hannah moved from simple descriptive explanations to linking the features noted by the students to the contribution of that adaptation to flight. Thus in the introductory lesson which highlighted the role of streamlining, Hannah focused not on the definition of the word (as did Lara) but on the shape of objects that students had identified as streamlined. She continually asked the students to work out "what makes it easier to go through air?". Later on in the same lesson, the children had an opportunity to share their thoughts regarding which parts of a bird are used to control its flight. The students offered one or two word suggestions such as tail feathers, shoulder muscles, wings, feathers. In each case, Hannah asked students to add what each particular feature was used for. Interestingly enough, while the students easily moved into relational explanations, they did not all transfer this level of complexity into their written work.

Both teachers in this study perceived the primary role of language in learning science as communicative; that in recognizing the labels used to describe, and occasionally in linking observations, the teachers believed they were inducting students in the language of science.

For Lara, language was not a tool for inquiry, but an addition to a communicative repertoire. In her talk, she emphasized "what we already know", rather than "how do we know this?". The classroom discourse was not oriented to generating explanations new to the learning community which could be seen as 'better' explanations than beginning explanations. Print resources were seen as sources of information, not as sources for ideas. Accepted relational explanations were validated by manipulated objects.

In Hannah's case, this approach was not appealing. Hannah acknowledged language as a tool for inquiry and, in other subject areas, she provided many opportunities for students to rehearse different forms and functions. But, in her science lessons, these were limited to extended periods to oral expression of students' ideas. While Hannah strived to develop an inquiry discourse, she had not recognized the process of constituting evidence through language and establishing relations and explanatory models. As in Lara's class, students were more inclined to utilize data to construct descriptive explanations than explanatory models.

Hannah assigned student task sheets but did not go over the student work, either on her own or through class discussion once the tasks were completed. This effectively disengaged her from student interpretation of the print resources and written text construction. In her reflections on the unit, she acknowledged that students "really have to write about it more". Many opportunities were provided in the units for both reading and writing, but Hannah took for granted these activities as being both natural and neutral. It was only when she came to read the flying animal reports and mark the tests that she realised the difficulties students were having in constructing explanations of flight and adaptations for flight.

The students in these classes were intrigued by the notions of flight, and displayed willingness to move from describing/comparing to explaining, arguing or persuading. As researchers, we believe that it may be possible to scaffold such a movement by acknowledging that students need many varied opportunities to use discursive tools in which links between relations, and the support for such links, are articulated. We see the role of the teacher being a guide to the consideration of new ideas, ways of investigating them, and ways of linking them to models and theories. In classroom communities, if students are to construct meaning for natural phenomena, they need to learn how to articulate the links between data, evidence, and explanations. Students need to recognise that competing explanations are possible, and that there are ways of discriminating between these alternatives. Furthermore, they need support and practice in constructing explanations that go beyond descriptions or loose lists of contingent conditions. Although the ability of children to work in the realm of theorizing has been questioned over the decades, extensive documentation of children's alternate conceptions illustrates that students do construct and bring explanatory models to the classroom. In school science, the challenge is to construct a pedagogical discourse which encourages the appropriation of tools for inquiry.

In the classrooms in this study, the teachers rarely entered into deliberative discussions about strategies for the construction of explanations and were reluctant to draw on textual resources as models of explanations, possibly fearing that students would defer to the authority of the text, rather than seeking to generate their own explanations (Palincsar & Magnusson, 2001). Clearly, a significant shift in the orientation of instructional discourse will be needed for teachers in primary science lessons to attempt to address this dimension of learning science. They will need to be convinced that language activities across a spectrum of modes are essential components in learning science. Teachers will also need access to a range of pedagogical strategies which initiate and support the appropriation of language for construction of explanatory models in school science. The way forward in uncertain.

In the introduction to this paper, we noted that the call to learn science through language use was being made more than a decade after similar exhortations. We are uncertain how the call will be heeded this time. New models of curriculum, instruction and assessment are required (Duschl & Ellenbogen, 2002). We draw on the case studies outlined above to probe a way forward, giving consideration to the constructs employed by the research community as well as to the shaping of instructional practices within the institutional context of schooling.

It has taken many years for literacy education researchers and science education researchers to begin the effort required to develop shared frameworks for dealing with the role of language in school science. The notion of literacy as "the control of discourses that go beyond whatever discourse we first acquired within our family" (Michaels & O'Connor, 1990, cited in Anderson, Holland & Palincsar, 1997) is only now beginning to be addressed by researchers in school science.

Today, we are in need of common constructs which address the acquisition of language forms for the purpose of constructing explanations within a discourse of science. That is, students need to be able to use language to transform data into evidence, and then to construct arguments which draw on this evidence to support claims in the form of explanatory models. In the past, such reasoning has been referred to generically as 'interpretation' of findings. "Making meaning" from the findings is another view of these processes. We need to understand how to scaffold student engagement in oral and written texts in the transformation of data to evidence to explanations.

Discriminating between information (data) and evidence would appear not to be an intuitive strategy, as indicated in these case studies as well as by the findings of Driver et al (1996). In what the latter researchers describe as phenomenon -based reasoning, individuals make no distinction between description and a phenomenon. However, the researchers suggest that this does not imply that individuals are not capable of making this distinction if scaffolded in doing so. The researchers make the point that theories and explanations are expressed in a different language from the language of observations (p.116). The notion of appropriating language forms to enable movement between discourses is prominent among researchers working from a sociocultural perspective, but this entails explication across disciplinary boundaries to clarify ways in which purpose is accomplished through various modes of language.

The difficulty of distinguishing between data and evidence and between data and claims being made is currently being investigated by science education researchers with a focus on argument and argumentation (Kuhn, 1993; Duschl & Ellenbogen, 2002; Osborne, Simon & Erduran, 2002). Duschl & Ellenbogen (2002) offer a scheme for using argumentation at different points in the development of explanation (theory). At the first decision making point, data are selected to be counted as evidence. At the second decision-making point, a selection of evidence is made to construct a model, and finally an explanatory model is selected.

However, even the construct of 'explanation' and its diverse categorizations of explanations highlights the difficulties facing researchers (Unsworth, 2000).

At this juncture, it is as if researchers are situated at a nexus of intersecting contexts for school science. And constructs derived from these contexts are competing for dominance. For example the notion of learning communities (Brown & Campione, 1994; Bereiter, Scardamalia, Cassells & Hewitt, 1997; Haneda & Wells, 2000) and its potential for learning through social contributions intersects with the attributes of science inquiry communities and their orientation to construction of explanatory models.

Discursive practices in learning communities

Primary teachers have little difficulty with the notion of the classroom as a learning community. Teachers of young students recognize the social nature of learning and provide opportunities for the sharing and inscribing of ideas. Haneda & Wells (2000) highlight collaborative contributions as essential to knowledge-building communities, in which students generate both oral and written texts as a result of deliberate and functional choices. These researchers note that deconstructing the organization of various genres - such as explanation - may assist students to see the functional significance of their characteristic features. However, Haneda & Wells suggest that it is equally important for students to reverse this process; given a particular topic, audience and purpose, to generate candidate generic structure and discuss which would be most effective, and why (p. 451). The range of discursive activities undoubtedly varies from one classroom to the next (as in the classrooms reported here), and subsequently the opportunities for students to appropriate discourses such as that of science inquiry. The uncertain aspect of interfacing the constructs of researchers working from a sociocultural perspective with the instructional practices of teachers is how to unpack the focus on language use across the curriculum in the professional community.

Classroom communities exist within the larger professional communities of schools and school jurisdictions. The practices of these institutional communities, in particular their curriculum and accountability practices, may be perceived as obstacles to giving close attention to classroom discourse. In jurisdictions with tightly prescribed learning expectations and testing situations, teachers are hesitant to break out of content-oriented instruction in favor of an orientation to language practices in their learning community. And views of literacy and language use are not aided by disciplinary orientations which neglect to acknowledge that students are being asked to move between multiple discourses in their studies.

Since the aim of science is to construct tentative explanations of natural phenomena, an inquiry approach to school science should encompass ways of talking, reading and writing such that students begin to appropriate the reasoning that transforms data into evidence supporting explanations. These will entail language which goes beyond the exchange of ideas, as seen in the classrooms reported here. The instructional discourse will entail conjecture, rather than certainty, as students develop explanations by proposing unseen mechanisms to account for phenomena or events. Efforts to introduce prospective primary teachers to the discursive practices of science through talking, reading and writing illustrate the challenges of confronting both authority and uncertainty in the same discourse (Smith & Anderson, 1999). And when the personal reservations are overlaid by institutional constraints, it is scarcely surprising that only the most confident of primary teachers will undertake to open the gates to an examination of how to construct explanatory models in school science.

Literacy practices in school science will need to be viewed not as acquisition of individual skills but as contributions to a collective practice in the classroom community. Within this collective practice, language is used to assess sources of evidence and evaluate knowledge claims (Smith & Anderson, 1999). The challenge for researchers and teachers is to develop shared ways of conceptualising the literacy practices of knowledge-building communities (Anderson, Holland & Palincsar, 1997; Bereiter, Scardamalia, Cassells & Hewitt, 1997; Haneda & Wells, 2000). In school science, a knowledge-building community will engage in deliberate argumentation or persuasion about explanatory models (for example, Osborne, Simon & Erduran, 2002). Some researchers have investigated the development of rubrics for constructing explanations or arguments. Although such frames, such as those put forward by Wellington and Osborne (2001), provide a structure on which to hang an explanation, they do not offer an audience or purpose. Furthermore, they do not locate the construction of the explanation within an inquiry context, demanding explication of ties to data and evidence.

Finally, it seems uncertain that teachers will consider the literacy practices of science communities to be of primary significance until the priorities and practices of the institutions of schooling do likewise.

1. Birds have a number of adaptations that enable them to counter the effects of the four forces that impact flight (thrust, drag, lift, gravity). The bones of birds contain hollow spaces which lessen overall weight and allow birds to generate sufficient lift to overcome the force of gravity. Birds have the ability to generate thrust through specific wing movements, and can manipulate their wings to increase the angle of attack in an airstream, thus developing lift. Their wings are curved in an airfoil shape and the movement of air over the wing surface during flight helps maintain lift. Both the body shape and the careful overlapping of feathers streamline the bird, thus reducing the effects of drag as it moves through the air. The wing and tail feathers can also be manipulated to increase drag when a bird wishes to slow down by decreasing thrust. Finally, the muscles of birds contribute to their ability to fly. In continuous flapping flight, a bird can generate about 10 times the mechanical power (per kilogram of muscle mass) that a human muscle mass can. During a long flight, bird pectoral muscles metabolize fats directly, whereas human muscles burn sugars. The nutritional value of fat is twice that of sugar. Hence birds' muscles are designed to not only generate, but sustain necessary thrust.

Swales, G. (1990) Genre Analysis. Cambridge: Cambridge University Press.