Cultural Tools and Apprenticeship: Learning in Science

By Richard Sohmer & Sarah Michaels, Jacob Hiatt School for Urban Education, Clark University, Worcester, MA USA [1]

Several years ago, in Alice Springs, Australia, one of us (Sohmer) gave a talk to educational researchers and teachers about the Investigators' Club, our after-school physics program for "at-risk" 7th and 8th graders. In his enthusiasm, Sohmer told them everything he could squeeze into his allotted (thirty-odd) minutes about the "I-Club" - the problem(s) it addresses, the sociocultural theory underlying its design, that it's an apprenticeship in doing science, that it "presumes intersubjectivity in order to attain intersubjectivity," and that it works, that kids graduate from the I-Club as self-identifying Science Investigators, knowing how to observe, think, and argue appropriately as well as knowing "stuff [about physics] that most college grads don't know." The responses of the Aussies in the question period that followed were gratifyingly positive. So when they asked "What impact is your Investigators' Club having on science education in the States?," they were astonished to hear his (chagrined, lame) reply, "None, yet - it's really only a research site, at present."

Later, back in the States, a middle-class mom (a professional transcriber who had just finished converting the audiotape of Sohmer's Alice Springs talk into a Word file) told him how thrilled she was to find out about the Investigators' Club, and asked him what chance there might be of "getting my kids into the Investigators' Club." All Sohmer could offer her was the same (equally lame) reply: "I'm sorry - it's only a research site, at present." She was keen to tell him about her kids, anyway, and the conversation continued, as follows:

Mom: See, my daughter was always great at science
She always got A's and she always loved science
til she got to the seventh grade.
And then, the first day,
she came home,
and said
"Guess what I found out at school today, Mommy -
I'm no good at science."

RS: God, that's horrible!

Mom: Yep.
And the next year,
my son went into the seventh grade
at the same school
and got the same teacher
and the same thing happened to him.

Sohmer and the mom talked a long time and shared their rage; her face got red and tears glistened in her eyes as she talked about the damage done to her son and daughter. Sohmer did what little he could do to help her. He assured her that it wasn't her kids' fault, that it was, of course, an out-and-out falsehood that her kids were "no good at science" - and that it wasn't her fault, either, that she couldn't protect them from the intellectual maiming they had suffered. He reassured her, for what comfort she could draw from the fact, that she and her kids were not the only ones damaged. The one big idea that school "science" in the US successfully teaches most of its participants is the very same blame-the-victim falsehood her daughter had been saddled with: "I'm no good at science."

Sohmer told her that it was known--that he and many others knew--that "science education" in the US is, on the whole, a shameful hoax. The TIMSS studies (1996, Schmidt et al., 1996) have clearly shown that even those US students who do extraordinarily well at the game learn little actual science. He told her, too, that he didn't know how to fix the situation, that "science education" in the US seemed to be a lumbering, dangerous Frankenstein that believed its own press and would destroy any who called its fundamental operations into question. Together, they agreed that the best they could do, for the nonce, would be (in whatever degree possible) to refuse to collude in the hoax perpetrated upon our children, to both refuse to reify and to actively contest the "no good at science" evaluations, and to make common cause with others by "going public" with our disagreements with school "science" as it currently exists.

Well, you may say, so what? Here, we stipulate, are some notable "so-whats":

1) Kids start out inherently and deeply interested in finding out "how the world works" - what the Greeks called

("phusis"), i.e. physics. It is inaccurate, and ultimately morally wrong, to characterize the present state of "science education" in the US as "ineffective," or simply "less effective" than that of other nations. Such characterizations function to obscure the much-needed recognition that there is in fact something fundamentally wrong, and instead lends credence to the continuing spate of reforms, standards, etc. directed towards an improvement at the edges of the status quo. In brief, the current model of "science education" in the US does real damage on a daily basis to the selves and minds of most of our children.

2) Apprenticeship is the appropriate model of genuine teaching-and-learning in the case of highly complex interactional and intellectual practices. The current model of "science education" in the US does not, however, even begin to approach apprenticeship. The bottleneck is the central economic fact that we cannot afford to replace the great majority of "science" teachers who are teaching "out of field." As a consequence, these teachers are only capable of "business as usual" - i.e. "textbook science," in which they are tasked with the pro forma evaluation of students' abilities to retain and regurgitate disconnected and essentially meaningless texts and algorithms.

3) The asymmetry of power between teacher and student can facilitate discourse acquisition (by, e.g., licensing legitimate peripheral participation (Lave & Wenger) and it can also be profoundly pernicious. Those who are members of a coherent discourse have the possibility of apprenticing others into it; those who are not, cannot. For a variety of (understandable) reasons, teachers sometimes falsely represent themselves as being members of a given discourse, with the inevitable result that classroom interactions which are putatively discourse-instructive are in fact discourse-incoherent. Students typically locate the cause of consistently incoherent interactions in themselves. Thus consistently incoherent interactions arising from misrepresentation on the part of the school and its representatives are not simply ineffective; they are noxiously effective in engendering apathy, self-disgust, and de facto incompetence in their students. ("Guess what I found out today at school, Mommy - I'm no good at science.")

4) The good news is that our experience in the I-Club argues that "out of field" teachers can be retooled as (apprenticing) members of the discourse they are charged with teaching, based on three principles:

Apprentices are legitimate members of the practice (Lave & Wenger, Rogoff);
No member of a complex practice "knows everything" (Polanyi, Herbst);
Knowledge is distributed across the "actants" (persons and mediational means) of the practice (Latour).

In what follows, we'll describe the Investigators' Club and the implications it has for reforming teaching and learning science in the United States - for both students and teachers.

Background information about the Investigators' Club Principles and Practices

With colleagues at Clark University and teachers from the Worcester Public Schools, we've been working on the Investigators' Club for the past 7 years (funded by two major grants from the Spencer Foundation). As alluded to above, the I-Club instantiates a set of pedagogical, philosophical, and theoretical commitments in an after-school science program for inner city middle school students. The Investigators' Club is thus both an experimental research site and a long-term research and development agenda. The I-Club "program" is a cross-disciplinary effort in that it integrates:

work on the architecture of intersubjectivity (traditionally examined within the fields of psychology and sociology);
sociolinguistic studies of participation and identity transformation;
practitioner research focusing on teacher-student, student-student, and whole-group discussion; and
an analysis (founded on Sohmer's twenty-plus years of experience in teaching and building houses) of teaching and learning as apprenticeship.

The Investigators' Club ("I-Club") program meets three times a week after school with students from a wide range of cultural and linguistic backgrounds, predominantly low SES students who are struggling and/or failing in school. Unlike traditional school science, the I-Club recruits students' everyday ways of speaking about the world - while gradually scaffolding them into the use of new discursive tools (new ways of giving scientific explanations and using representational tools). In this program, the activities ("tasks" or "demos") are designed to promote active theorizing, prediction, and argument about puzzling physical phenomena, often called "discrepant events." In the process, having a well-argued theory is the name of the game. If a student's prediction, or theory, or both are in the end disconfirmed by the evidence, that is OK: the job of the scientist is to make cogent predictions and theories so that they may be cogently disconfirmed. The goal is to make one's claim as explicit and persuasive as possible. Everyone benefits from seeing the (ultimately) best theory in the field of contesting, less effective theories, and everyone can appropriate the results for their own use in the next task.

The Investigators' Club program presumes that students come to class with very well developed theories of how the world works. Students are, in our view, already successful investigators of the physical world - they know how to jump out of the way of an on-coming bus, transfer liquids, move heavy or clumsy objects around, deal with friction and force, etc.; they have (implicit) theories of invisible, underlying forces (suction, heat, pressure, gravity) in their environments. In contrast, what we call "Textbook Science" deprecates students' already-existing knowledge. At one extreme, the existence, extent, complexity, and utility of students' knowledge are simply not acknowledged. At another extreme, students' knowledge is framed as comprised of baleful misconceptions which are to be removed and replaced - like getting one's teeth swapped out for dentures - by canonical counterparts.

In the I-Club environment, heterogeneity of students' experience and cultural background is a valuable resource. When the group evaluates competing theories in their most persuasive forms, in the shared context of the demo at hand, cognitive growth in the form of movement towards more effective (and canonical) ways of seeing and talking is self-motivated and self-enhancing, grounded in individual and collective experience (rather than a concern for the "right answer"). The teacher also provides a new set of "explanatory tools" (often in the form of analogies and narratives) that help students re-see the world.

Our perspective on learning as apprenticeship entails the centrality of cultural tools:

Tools extend mind beyond the skin.
Tools transform the activity and the actor.
Tools and toolkits are "always already there" - i.e., access and use of tools are regulated (often zealously) by the Discourses to which they belong.
A crucial part of the replication of a valued Discourse is the ongoing apprenticeship of members into the coherent use of its tools.
Meaningful activity is accomplished when individuals both act with appropriate cultural tools and are recognized as doing so by other members of the Discourse.

Over seven years of the Investigators Club practice we have (so far) identified three types of cultural tools in the I-Club toolkit:

Demos -- create shared theorizable situations that can be argued about productively when physically present, as well as referred to, later, as benchmark experiences and sources of warrants.
Explanatory tools -- provide ways of thinking about and perceiving situations in new and productive ways.
Activity structures -- influence the nature of talk, position students as active practitioners of the Discourse, tied to specific content provided by the demos.

We will discuss each in turn.

1) I-Club "demos" create shared experiences and shared "contexts in the mind"

The demonstrations at the center of I-Club whole group ("Circle Up Time") discussions are not simply "science demos" of the traditional sort, observed by students and performed and explained by the teacher. These demos, by design, use commonplace, everyday materials (a jar with balloons, empty gasoline cans, candles, soda bottles, cups and water), and often involve something a bit "dangerous" or tactilely stimulating like flames, imploding cans, karate-chopping hard-boiled eggs, breaking yardsticks, opportunities to be splashed with water, siphons, water levels, and often making things to solve problems. [2] Each demo builds on students' embodied experience and highly developed physical intuitions but constitutes a counterintuitive, "discrepant event."

When [student] pumps15 pumpstrokes of air into this volleyball [using a bicycle pump and inflating needle], will the volleyball weigh more, less, or stay the same?
When I fill this class with water, put a piece of paper over the top and then turn the glass of water with the paper covering it upside down, the water doesn't come out. Why not? [Later - after lots of discussion of the "ocean of air" surrounding us, the air pushing with 14.7 pounds ("like the weight of two gallons of milk" ) of force on every square inch of everything - How tall do you think this upside-down glass full of water could be - and the water still not fall out? How could we find out?]

In designing these demos for the Investigators' Club, Sohmer (the original lead-teacher of the I-Club) drew from many sources-countless books of science experiments, Arons' classic textbook, the work of noted US physics teacher Jim Minstrell, even Japanese science casebooks. [3] From these sources, Sohmer devised a linked set of demos about the topic under investigation which everyone experienced together. I-Club members often assist with the setup and operation of the demos, and consistently report their interest in and satisfaction with Investigators' Club activities, characterizing them as "hands-on" (as opposed to text-oriented). It is remarkable then, that the demos are overwhelmingly not "hands-on" - at least not in the progressivist, anything goes - "Let a thousand flowers bloom" - sense. The demos are designed to focus everyone's attention on a set of key theorizable "situations."

uses ordinary materials --- so that each student believes that s/he can anticipate the outcome.
has more than one imaginable outcome --- so that students predict and argue for different outcomes.
is (in some aspect) a "discrepant event" --- i.e. produces effect(s) contrary to those which most students' implicit theories predict.
is poised to run, but is not run until after students fully engage each other's differing predictions, arguments, and evidence.
locates "the right answer" in the world (the demo's outcome), not in the teacher.

A number of "demos" are presented in the photos here.

Each demo constitutes a challenge to the student's present understanding of the world, an intriguing puzzle which is nonetheless theorizable because it is (except for its discrepant outcome) located within his/her existing experience base and language competence. These demos become the source of shared experiences in the group but push for divergence in theories or positions; this sparks extended animated discussion, conflicting viewpoints, and arguments (bolstered with evidence). What these carefully selected "public" events accomplish is the creation of a set of shared situations to think and theorize with, a set of shared "contexts in the mind," to use Courtney Cazden's phrase (1992). They regiment experience and constrain the search space for the group. It becomes a way of managing diversity and creating common ground, amidst a wealth of divergent experiences in the world and a wealth of discursive resources, among an ethnically and linguistically diverse group of students.

The demos are designed to give the Investigators both key phenomenal experiences in the world coupled with experiences of developing and arguing over and for theories. Below is one example of the talk prompted by an I-Club demo. This particular demonstration involved an empty metal can (gallon-size, the kind that paint thinner comes in) heated up over a Bunsen burner. At the outset, Sohmer handed the "empty" can over to the Investigators for examination. As the can was passing around the circle, Sohmer asked them for their observations. It's empty. It's metal. I can hear the ocean in it. The lid screws on. Upon request, an Investigator poured a "very small amount of water" (approximately a tablespoon) into the metal can. Sohmer lit the Bunsen burner and placed the can just above the flame, on a ring stand. The water inside the can could be heard coming to a boil. Then - with steam visibly gushing from its spout - Sohmer removed the can from the flame and held it (with oven mitts) while an undergraduate research assistant (who usually manned the video camera) quickly screwed the lid on. Sohmer then set the can on a table in everyone's sight. The group sits there, waiting, watching. The burner gets turned off. After all that drama, what seems to be happening is, at first, exactly� nothing. Suddenly the can emits a creak, a kind of tinny-sounding ping, and then another. Everyone is riveted to their seats, eyes on the can. More pings, then a big ping: the large sides of the can snap inward. Over the next two minutes, the metal can continues to warp, bend, crush, seeming to be "sucked in on itself". [4] In these transcripts, we refer to Sohmer as Richard or, simply, R.

Richard: What do you think happened with the can? // What happened? // Why--why did it do that? //
Mark: It like / it looks / kind of sped up the process of decomposing / and the-deteriorating /
R: Uh huh / wh-what-what of-- what's deteriorating / what's decomposing //
Mark: um decomposing / deteriorating means to fade away // first it starts to / get smaller / [then little pie--] it starts to / rust and get thinner / inside / and just rust away //
R: Oh so you're saying that the inside of the can has been--is rusting away //
Mark: It eats right through it //
R: OK // Who else? // (invokes gender rule) / so we had one guy and now we gotta hear from one girl //
Anna: What was the purpose of the water? // and / how come / when it went and it was heated up / [...] I don't understand what happened //
R: That's what we're trying to figure out here // but what's your theory? // I mean I could tell you what I think // that would be kind of / no fun //
Anna: It can't be ru:sting // it's just--there--there's just / water in there //
Tyrone: (overlapping)
R: [invoking the speaker has the floor rule] Wait-wait // let her say it // let her get her chance to say it //
Anna: the water just / sort of evaporated / a:nd when you cover the top / there was so much heat inside it / it's probably just / melting or / sucking it in [..]
R: so what -- so you say -- it seems to me I hear you say it's-it's not rusting // and something else // what else // what is it then? //
Anna: It's hot in there // [...]
R: You said-- you thought it may have something to do with melting //
Anna: The water / wanted to [...]
R: I'm not so much interested in getting right answers as I am to try and figure how to get at / get / get close to the right answer //
Maria: The air was like too hot / so it [...] trappin' the air in there //
R: So.. (has to do with air being hot)
Tyrone: Um / 'cause I was reading this book // well it didn't say it but it showed a picture / of like / you put um / you fill a balloon um with air / [R: Yeah] and you put it in something hot / and the um / water molecules expand / and um they / make the [..]balloon bigger / and then when you put in ice / or colder temperature / it shrinks down / and all the / uh 'cause all the water molecules like / contract / and the balloon gets / smaller / so maybe that would be the same with the tin can / [R. nods yes] Nod yes I'm right / nod [..]
R: I'm-I'm nodding yes because that sounds like a theory // but I'm not nodding yes because that's the right theory //
R: What happened? // ... to make sure we all saw the same thing //

This example highlights many of the common features of the talk during demo-related activities - Sohmer's revoicing of students' contributions, invocation of selected turn taking norms (the gender rule and current speaker has the right to the floor until finished); as well as the students' use of resources (previous experience, intuitive, everyday notions of rusting, melting; the use of big, scientific-sounding words, such as "deterioration," etc.) they brought with them from their everyday lives and previous school experience.

We want to be clear that this kind of "demo talk" is quite different from small group hands-on science (Cohen, 1997; Ballenger 1996), more characteristic of progressive pedagogies, where a divergent set of current and past experience are recruited as resources. It is different in the regimentation created by the use of these shared situations, and by the voting on predicted outcomes. In this case, everyone focuses on the same thing, but brings divergent positions, predictions, and theories to the table.

2) Explanatory tools -- ways of thinking about situations in new ways (via models, analogies, narratives). Carefully designed explanatory tools recast the situation under investigation, making structural regularities and causal mechanisms explicit. These tools work to help students re-see or retheorize physical situations the world. In this way, they engender competence and a sense of efficacy in the students as they attempt to explaining complex scientific phenomena. What is required for an explanatory tool to become a good, useful tool? The tool: omust make use of accessible, familiar notions; ohas to achieve perceptible results for the user; omust strike the user as giving better results than their current tool.

However, we have found that it's not enough to expose students to powerful conceptual and explanatory tools. They have to know when and how to use them. In order for tools to be successful, novice users:

have to have the identity and the right to pick up the tools;
need modeling and practice in using the tools;
need to learn when to use them;
have to develop the ability to signal to others that they are using the right tool at the right time in the right way.

Science is a domain in which non-narrative modes of discourse tend to be privileged in analyses of talk and text (Halliday & Martin, 1993; Lemke, 1990). In the Investigators' Club, we have found that narrative tools play a far more important role in science apprenticeships than the literature would suggest. In particular, participants in the Investigators' Club have been markedly successful in using a specific narrative tool - the "Two Puppies" story - to understand and explain the physics of air pressure. [5]In analyzing the use of tools in the service of scientific explanation in the Investigators' Club, we ask: What are the tools which help students see, speak, and reason in new and powerful ways? How do you get middle school students who hate school, and who (by and large) are failing at school to pick up these tools willingly and use them effectively? When a tool - especially a narrative tool, like the Two Puppies story - works extraordinarily well, we are especially curious. We observe that it works - but how does it work?

We have explored the affordances that the "Two Puppies" story has provided over non-narrative alternatives. In what follows, we provide examples of the tool being used (taken up and practiced) by students, and examine the ways that students' use of the tool transforms both the user and the accomplishment of scientific explanation. We theorize that this tool serves as a bridging metaphor, mediating between concepts and actual situations in the world, helping students re-see the world and their place in it. [6]

Background information about the science at issue and the narrative tool, the "Two Puppies" story.

An important feature of explanation and argumentation in physics is that intentionality, desires, and emotions are not to be ascribed to physical agents or processes. Beginners mark themselves as not being in the Discourse (Gee, 1996) of physics by their na�ve use of anthropomorphic arguments: "The fire needed oxygen so it had to get the egg out of the way," or "The smoke rose up in the bottle because it wanted to get out." Even the scientific sounding (and much resorted to) pseudo-law, "Nature abhors a vacuum," is an example of this kind of reasoning - based on an understanding of physics as being like the human world in which emotions, promises, irony, desires, lies, and goals play a major role in our understanding and explanation of events. Because explanations in physics do not trade in these everyday, implicitly narrative forms of causality and explanation, one might expect that the process of inducting or apprenticing someone into the Discourse of physics would privilege the use of tools that are themselves decontextualized and non-narrative in nature.

In the "Two Puppies" story [7] , the "puppies" referred to are mythical or fictional beings - "Air-puppies" - combining some of the properties of real, live puppies with the behavioral characteristics of air molecules. The air-puppies are the bumbling (mindless) agents in a modifiable story with a particular setting (always including two rooms separated by a moveable wall-on-wheels), participating in a series of events, always resulting in some kind of lawful effect - that is, the wall moves as it must, given the air-puppies' opposing impacts upon both sides.

We typically introduce the Air-puppies story to the kids (in a 10- to 20-minute session) by telling them the basic story, followed - always - by several variations.[8] As the story progresses, the situation and changes in it are illustrated with simple, freehand drawings (on whiteboard, paper, or chalkboard). We begin by asking the kids to imagine a big room divided into two smaller rooms by a wall on frictionless wheels (like roller skates). In each of the rooms on either side of the wall-on-wheels there are air-puppies - initially, (equal numbers and kinds of air-puppies - mindlessly bumbling around. (Figure 1 below shows a top-down view of the situation.) The dividing wall-on-wheels moves [9] whenever a puppy bumps into it (not intentionally, just mindlessly moving around). As the puppies bumble around and mindlessly bump into things (all the walls and each other), "What," we ask the kids, "will happen to the wall?" Even at this point, in this first session, one or more kids will confidently "read" the situation to predict that "the wall [on wheels] will stay in the same place."

Figure 1. Click here. The view from above of the beginning of the "Two Puppies" story. In this version of the story there are equal (numbers and kinds of) airpuppies on each side of the wall-on-wheels.

Once the scenario in Figure 1 is set in motion the wall-on-wheels (henceforth simply, as the kids say it, "the wall") is pushed a little bit to one side or the other each time a puppy bumps into it. Because the wall gets, on average, the same number and kind of bumps from each side, the wall stays over time in approximately the same place, oscillating about the centerline (Figure 2).

Figure 2. Click here. With equal (numbers and kinds of) airpuppies on each side, the wall-on-wheels is continually bumped from side to side. The net impact of the puppies on one side of "the wall" (the wall-on-wheels) is, on average, equal to the net impact of the puppies on the other side, making the wall oscillate about the centerline.

But if, say, some of the puppies on the right side of the wall leave the room (by a door), what will happen? Figure 3 illustrates this situation.

Figure 3. Click here. View from times 1, 2, and 3. As airpuppies in the right room "bumble" out the open door, there are fewer and fewer airpuppy impacts from the right upon the wall-on-wheels. Increasingly unopposed airpuppy impacts fron the left push the wall away - to the left.

In this case, over time, "puppy bumps" on the left side are less and less opposed by puppy bumps on the right side, with the result that the wall-on-wheels is pushed to the right.

Most people see, use, and accept "suction" as a perfectly adequate explanation of ordinary actions like using an Electrolux to clean a carpet or drinking a milkshake through a straw. An ordinary person who doesn't know much physics sees sucking (or, what sounds more scientific, a "vacuum") at work when they see a person drinking a milkshake through a straw. A physicist, in contrast, sees pushing. The actual forces of pulling and pushing are both invisible, but practitioners of physics see pushing: atmospheric pressure pushing the milkshake up into the straw. Here's what Andy diSessa, physicist and science educator, writes on "suckers and sucking:"

...Initially the overt agent, the person sucking on the straw, is viewed...as the direct cause of the liquid's motion. ... Later in development, at the level of a physicist's informal description, an invisible, inanimate agent (i.e., air pressure pushing down on the liquid in the glass) becomes the direct cause of motion; it pushes the water up the straw. This happens when the balance of pressures (pressure on the water in the glass balances pressure on the top of the liquid inside the straw) is broken by the overt agent - the person sucking. It is notable that, at this stage, in contrast to earlier stages, a vacuum or partial vacuum can in principle never pull (suck) at all. Sucking is entirely gone as an explanatory primitive. In a bit more detail, the new analysis is: The person changes the geometry of the situation, for example, by increasing the volume inside his or her mouth in which a fixed amount of air exists. This results in decreased pressure on the liquid on the straw-and-mouth side of the system. A decreased pressure on the inside of the straw and an undiminished atmospheric pressure on the other side cause the water to move. (diSessa, 1993, pp. 144-145).

The initial invention and use of the Two Puppies story stems from Sohmer's observation that, in practice, novice physics learners simply do not retire or replace their good old "suction" as an explanatory tool after (repeated) reading or instruction in the details of a canonical account (like diSessa's) of air pressure. By contrast, these same novices do take on the Two Puppies story - which encodes the canonical explanation in a narrative - to successfully "re-see" the physics of air pressure.

Example 1: Students guided in the use of the "Two Puppies" story As an example of I-Club students working with this narrative tool, we draw on a segment of talk taken from the 1997 I-Club. Over a period of days, and through literally hundreds of turns at reasoning, Sohmer pushed and prodded and scaffolded the students to work with the Two Puppies story, unpack it, model it on paper, and think about what it bought them. He also led them to talk about why scientists talk about air pressure the way they do and how it's different from the way everyday talk works. This was always done in the context of applying the current topic of talk to old and new demonstrations, some present, some merely shared contexts of the mind (Cazden, 1992).

The following example [10] shows one instance of a lengthy process of Sohmer scaffolding the students in using the narrative tool. This discussion took place during the 5th week of the I-Club, long after the kids had first been introduced to the Two-Puppies tool itself (which happened on the 4th day of the I-Club). This discussion occurred in the context of an experiment that involved a metal one-gallon can, full of water, elevated on a stand 8 feet high off the ground. A tube attached to the can was, at the outset, clamped shut so that no water could run out. The bottom end of the tube lead into a large plastic bucket on the floor. (Figure 4 is a sketch of the set-up for this experiment.)

Figure 4. Click here. The can and long tube demo.

Sohmer asked the Investigators to predict - using the Two Puppies story in their account - what was going to happen when the clamp at the bottom end of the 8-foot long tube was opened. But, before they began, he once again practiced the Two Puppies story with them.

R: We're gonna run this experiment that the uh / that the girls set up / and- and then I'm gonna ask you to predict / what (...) / and uh and not just predict what's gonna happen / but (...) explain it / and explain your prediction using the Two Puppies theory // even if your- even your prediction is wrong / you could gotta- if you use it here / ingenious use of the Two Puppies theories / you get uh glory / uh of sorts // Cause it takes a while to figure out how to use this tool okay? / this- this explanation you might- / might use it wrong // But um / but so I'm gonna / point to you / and ask you a qui- / few questions about the Two Puppies theory / remember that there's the- there's a / the original version of this / it's a- it's what we call a metaphor / it's not really true / we're not talking puppies / we're talking about air molecules right? / really but / it's- it's convenient to talk about puppies / ...(reviews the story with the rooms, wall and 100 puppies)... Now what I want just to / just to check out how we're- how we we're doing about /applying this- this uh / tool / the Two Puppies metaphor // I- I'm look / I'm now I'm entertaining suggestions as to how we could -- / let's say here's / let's say this is the wall right here / And this will be room A / and this will be room B / we know / we want this (...) / this wall / to move that way / towards room / th-this is room A / this is room B // We want to get / the wall to move towards room A // There's about seven different ways we could do this // Just give me one way //
Mark: Push it //
R: Okay // what- we can't get in the room and push it cause only puppies are allowed in there // so how- /
Mark: Puppies push it //
R: So why the- so pup- so pup- while it tries to push on this side / but puppies are also pushing on this side //
Mark: Put all the puppies on this side //
Phil: How are you going to get 'em through the wall?
Mark: Shut up //
R: Okay no / no I'm sorry / he's got a (point) / we're gonna- gotta expand our version of this // gonna have in- in each room / uh into each room / might have a door / that you can open or shut / okay? // so you can put more puppies in or you could take some puppies out // Okay? //
Tyrone: Excite the puppies that are in the room / that- if you want to do all of them / that way you took the puppies that are useful //
R: Excite the puppies in this room / yeah how? // and you could excite them by? //
Tyrone: Heating 'em up.
R: Heating 'em up right // Here's where our metaphor crosses over into the air molecule / we're- we're fighting that / if you heat the puppies up they're gonna get excited and run around / if you cool down / they're gonna / calm down not run around as much // So my [...] / heat the 100 puppies up here // They're gonna run around more / they're gonna hit the wall more often / so the wall's gonna go this way // It's gonna go this- okay so what's another way we can move the wall this way / if we heat the puppies up /(...) // What's another way Phil? //
Phil: Uh / yell food //
R: Okay so / yelling food would- would do what? / get 'em excited? //
Phil: If they was / like / hungry / yeah //
R: Okay so but that- that's where this- this thing crosses over into where we're really talking about / air / air molecules uh so uh // the puppies is just a metaphor // so we've got one way of exciting them which is to heat them up // get them to move around faster / in fact heat just means getting them to move around faster // but we're gonna heat them up by putting heat in the floor // so uh you have another way? // What's another way? // [?: Uh] I'm sorry Anna / we're I'm not [?many of them] //
Anna: Because like um / cool them down on one side / and on the other side / the puppies would still be banging / on the thing / so [?it could get them to move] / on the thing //
Mark: Um / you can like-
R: Wait- wait did you get that one? // [?: Uh huh] You cool down the puppies on this side / on the- on the side you want it to move towards / and the- and the- the puppies on the other side are still / running around normally / so they're gonna bang the wall and [what's another?] //
Mark: You taunt 'em by um / you put like / a slab of meat in like a container / and poke holes in it / so they can smell it / and you hang it from the ceiling /

This example shows how overt focusing on the narrative as tool involved both guided practice and explicit instruction. Sohmer reminds them in lecture format (much of this turn is excerpted) and then gets the students to work with it themselves. At this point, through positive and negative evaluations, revoicings ("Wait-wait did you get that one? ...") and follow-up questions ("But how..."), he guides them to understand and work with the story in his way. He combines elements of an authoritative stance (there are right and wrong, better and worse ways to say it) with occasional revoicing of students' contributions, but only when they approximate what he's looking for. In this way, a disparate set of conceptions and "voices" is being recognized but then shaped, regimented publicly, to sound more like him. This excerpt also shows just how complex a process this is. Over time, the students had to come to see what aspects of the story they could retain (getting the puppies excited by heating or cooling) and what aspects were irrelevant and distracting (getting puppies excited by offering them food "if they were hungry").

Over many, many instances of practice, combined with overt instruction (repeated reviewing of the story - in talk, on chart paper, and later in computer generated models), the students were guided, sometimes in a completely explicit manner, to theorize with a new set of discursive and visual/representational tools. Most striking was the evolving (and eventually regular) use of the term "air-puppies" in the students' explanations. This usage served as a transitional conception for air molecules, as the Two Puppies story got taken up into their theorizing and into diagrams as a kind of inscription (using Latour's 1986 term). The air-puppies were variously represented as dots on paper, numbers inscribed in a room, etc., and increasingly were used with only the relevant characteristics of molecules (constant motion, bumping force, no intentions, no desires, no sense of smell). [11] Similarly, "Two Puppies" became an I-Club technical term - used by Sohmer and the I-Club members alike. It indexed the preferred air pressure tool, a deflating abstraction (in Bruno Latour's sense) about molecules, whereby two sets of competing puppies could be used to represent, understand, explain, and predict complex physical events in the world.

View two photographs of students representing puppies in their drawn explanations of air pressure phenomena. In one case, the puppies are represented as dots; in another case, they are represented as a number and temperature (e.g., 100 puppies, calmed down) relative to puppies on the other side of the wall.

Example 2: The Spud Gun
The following example shows the way the Two Puppies tool is taken up by the students, with support from the task, the teacher, and fellow investigators. The episode is taken from Day 6 of a different I-Club (Fall, 2000). It is only the second week of the I-Club and the norms for "Circle Up time," the group discussion activity supporting argument and peer critique, are still being developed. Nonetheless, the group has already discussed, theorized about, and argued about a number of different air pressure demos. They have been introduced to the Two Puppies story and the notion that air-puppies represent air molecules.

During this particular "Circle Up time," the students are asked to give an explanation of any of the many air pressure demo they have seen thus far, giving their best "scientific" explanation of it. In addition, the game requires that "you've got to be clear enough that someone listening to your explanation over the telephone would understand it." If the student's explanation is deemed acceptable by the other students, the presenter gets to shoot the "spud-gun" (itself an air-pressure demo), trying to shoot down as many film canisters, stacked in a pyramid, as they can in 30 seconds.

Although the game-like structure of this activity and the students' excitement about shooting the spud-gun disguise it, the teacher has several pedagogical purposes for getting the students to present and for getting the audience to critique the presenter. In the teacher's words:

First, presenting to their peers helps prepare them for public presentations and questions from an audience, such as science fairs or teaching younger students [both valued activities the students work toward in the I-Club]. Second, giving presentations to the group helps both presenter and audience take on the identity of an expert - an expert investigator, far different from their identities in school, where many of these students are failing. The student presenting stands up as an expert --- even if she has to try a couple of times to get it just right. The feedback, from peers, and from me, provides the presenter with guidance and practice so that she can improve as an explainer. At the same time, the members of the audience are positioned as "expert enough" to critique the presenter. I've put them in the position of the coach or teacher. Now, in this excerpt, they're just starting out, but because it's a task shared by the entire group, no one of them has to do it all. I've shown them I know they can do it, but, at the same time, I haven't given them a blank check. They have to do my job credibly in order to keep playing it, and I'm right there to step in to assist, if need be, either the presenter or her critics.

In this example, Daheesha selects the spud-gun, itself, as the air pressure demo she's going to explain.

Daheesha's turn lasts 5 minutes. In summary, here's what happens. Daheesha is called on. She stands up and says she's going to explain how the spud gun works. She first gives a step-by-step and somewhat stilted procedural account (with some help from her peers in naming the parts of the spud gun). She explains, in a sense "how you work it."

The other I-Club members are not satisfied with her account. They tell her she's got to: explain why the spud gun works, be more scientific by using the Two Puppies. Daheesha demands a second chance, saying, "this time I'll amaze you guys." On her second try, she shifts from procedural to explanatory mode - saying, "now how it works�" She gives two explanations in succession. She begins her first why-explanation by saying that the spud gun works "because there's a little spring in it." Laughter erupts and someone quickly points out that Richard had pointedly taken out the spring (view here) at the outset of the activity. Without missing a beat, Daheesha goes on to her second why-explanation saying that there's air inside the chamber and that's what makes the potato fall out.

Her peers are still not satisfied and again tell her she's got to use the Two Puppies, explain about the wall, "how when you push [the piston] the wall comes up and squishes [the puppies] in so they run into the wall more often and the potato'll go out." Daheesha presents once again, this time using the Two Puppies story as an explanatory tool. Interestingly, in this third account, as she moves into narrative mode with the puppies, she changes her discourse style dramatically. She shifts into Black English Vernacular syntax and prosody in giving a fully performed narrative (Hymes, 1981 calls this kind of shift a "breakthrough into performance"), complete with dialogue and animated hand gestures.

Daheesha: right now the air-puppies are havin' space // �now that / now I put it in [looks down and pushes the red piece into the black piece] / and they don't have no space so they-so they're like [high pitch, hands moving wildly] Oh let's get out of here // So they .. push outta this hole thing // and they all sss-[flying motion] ... [push out the potato]

The students register the improvement in her account ("Better. Better.") and vote that her account is good enough to warrant a turn at shooting the spud gun.

These examples of students using the Two Puppies tool are representative of something we've experienced in the I-Club practice hundreds of times and with a number of other audiences (ranging from first graders to audiences of school superintendents, public school teachers, and principals). In each case the Two Puppies story has enabled members of these various audiences to understand and explain, in effect, to "see" air pressure phenomena which were minutes before completely opaque to them. And, it's worth noting that until we developed the Two Puppies story, we had nowhere near this kind or degree of success in moving people to an understanding of the physics of air pressure. More positively, the success that people have had using this tool has made it a model for us in trying to design more tools like it.

What is it about the Two Puppies story that is so powerful, accessible, and transformative? Could the success of this tool be indicative of something fundamental about the way people think? Bruner (1986) and many others (Gee, 1986, 1996; Morrison, 1994; Hymes, 1981) have argued that narrative is a fundamental means by which we make sense of the world and our place in it. In a related but different vein, scholars interested in intrinsic motivation (Lepper et al., 1982; Cordova & Lepper, 1996) have claimed that there are several teaching strategies which positively affect intrinsic motivation and the process of learning academic content. One is what they refer to as "contextualization." "Presenting learning activities, even those involving abstract operations, in meaningful contexts of some inherent appeal to children should have significant beneficial effects on children's intrinsic motivation and learning" (Cordova & Lepper, 1996, p. 715).

The strategy of embedding cognitive activity within highly appealing (game-like or personalized) contexts may have something to do with the success of the two Two Puppies story. Kids and adults know about and like puppies. They don't feel intimidated by them and know that in some way they are not predictable. You can't tell puppies what to do so the idea that they are mindless but in constant motion is a graspable, familiar idea. There is enough overlap between the characteristics of puppies and the characteristics of air molecules that puppies have the potential to serve as a model or metaphor for air molecules. And of course, as the first transcript shows, there are some characteristics of puppies that do not overlap with those of air molecules and this is something that needs to be addressed and worked on. As we've shown, the kids are able to differentiate and rule out aspects of puppies that don't apply to air molecules. They learn readily that these are not puppies but rather "air" puppies.

Part of what is going on here surely relates to the claims of Cordova and Lepper (1986). The notion of Puppies makes the tool initially appealing and accessible. But this cannot be the entire explanation. The cuteness of puppies soon recedes as the focus shifts to air-puppies and the two rooms divided by a wall on frictionless wheels. These air-puppies are not cute or cuddly. They never stop moving to sleep. They don't eat or poop. No one has ever seen them and they are represented by dots.

What we think is more crucial than the mere familiarity of puppies is the dramatic contest (in dramaturgy, the "agon"), in which they are embedded. The groups of puppies, divided across two rooms, are always both separated and kept in ceaseless contact by the movements of the wall - which are always the simple, lawful consequences of (aggregated) movements of the individual, lawful movements of the individual puppies. The situation is always a battle between two opposing camps, or forces. Thus what is carried by the Two Puppies story is (at least):

the opposing camps of air-puppies
the wall that separates them and moves in response to their impacts;
the on-going, continuous movement of the puppies on both sides of the wall;
the non-intentionality of all these physical processes;
the indication (symbolized by the external walls) that everything outside this story world is irrelevant to the motion of the puppies and the wall;
and hence the law-like predictability of the movement of the wall as outcome.

This analytic list of its characteristics doesn't account for the efficacy of the story in use. Simply telling people these points (or their non-narrative equivalents) doesn't work [12]; telling them the story does. (The Investigators are able to represent, understand, explain, and predict complex physical events in the world, and to teach others how to do all of that - using the Two Puppies story.) In some (to date unexplained, but welcome) way, the relevant concepts, their relationships, and how they work together are carried by the story of (and ultimately indexed by the name of) the air-puppies. And that is why "puppies" are not a simple substitution for the word "molecules." Air-puppies carry with them the setting (rooms and wall), the rules of engagement, and the plot (the contest). Molecules do not.

Another important point in its favor is that the Two Puppies story does an effective job of replacing the black box notion of a vacuum in Investigators' explanations and accounts of air pressure phenomena. In any given case of the Two Puppies story, it is the presence of puppies (the aggregate of the numbers and forces of puppy bumps) - not the absence of puppies - that accounts for the wall's motion.

Even though this story is abstract, represented by lines and dots on chart paper or a whiteboard, and not something anyone has ever physically experienced, this story becomes a kind of new embodied knowledge for the students. They can think with the story, and run the contest in their minds - once, of course, they have located the wall and the two opposing camps of puppies. In fact, it's quite easy for staff members of the I-Club to tell when students have taken on the Two Puppies story as a way of seeing into new air pressure situations (a new demo, for example). They start out by saying something like, "OK well where's the wall here?" What they are trying to locate in this question is the point of contest -the wall - between the two opposing camps of air-puppies. Having located the wall, they can specify the states and conditions of each of the two sets of air-puppies and the size and characteristics of their respective "rooms," and the probable effect of any transformations undergone by either or both sets of air-puppies. It appears that air-puppies becomes a kind of abstraction (in Latour's terms, a deflating inscription), but unlike molecules, the term inextricably carries with it a set of abstract relationships and changes in time and space. The story carried by the term "puppies" automatically indexes setting, contest, and rules of engagement.

3) Activity structures - talk/task amalgams that organize participation, positioning of participants, roles and identities, linked with content.

Investigators' Club tasks (incorporating both demos and explanatory tools) are embedded in a set of participant structures and expectations which model the way scientists actually talk, think, and act. Students accomplish both identity work and cognitive work in the practice of these activities - and it is the students who are doing the science. In analyzing the stable talk routines of the I-Club, we have identified particular academic and discursive properties, particular constellation of intellectual roles and talk moves. One talk format is a whole group discussion in which a student presents an explanation and is critiqued by his or her peers (as in the spud-gun example above). Another recurring format is what we have come to call a "position-driven discussion." [13]

In "position-driven discussion," students focus on a single problem or situation. Each demonstration lends itself to extensive pre-demonstration prediction and theorizing, and post-demonstration theorizing and explanation. The demo usually has multiple possible outcomes or explanations, any one of which can be argued for on the basis of one's knowledge or experiences, and no one of which is obviously correct to the students. Students take positions and attempt to muster the best arguments and evidence they can in support of their position. Sometimes, informal votes are taken to see where the students stand with respect to one another, followed by more opportunities for students to change their mind, argue, and revote. In position-driven discussions everyone is focused on the same phenomenon but encouraged (indeed required) to commit to one position or another and to argue for their respective predictions or theories. Everyone is also free to change their position on the basis of another's evidence or arguments - typically with the proviso that you say (as specifically as possible) what it is in the other's position that you find useful or persuasive. This generates extended animated discussion, conflicting viewpoints, and arguments (bolstered with evidence). [14] This structure supports, in Courtney Cazden's phrase, "performance before competence" (Cazden 1983). Everyone is in the Discourse, before they have mastered the tools of the Discourse.

The teacher's job is not to provide "right answers." Put more positively, one of the I-Club teacher's major concerns is to avoid shutting down the discussion by prematurely "telegraphing" (indicating in any way) which theory is closest to being canonically correct. Rather, in position-driven discussions, the teacher typically scaffolds students by "revoicing" their contributions and pushing for clarification, so that everyone has access to everyone else's reasoning. The teacher might say, "OK, so let me see if I've got your theory right. You're saying that the volleyball will weigh less when I put more air into it because balloons are lighter when full of air?" Having a good "sayable" theory (conjecture, or position) is more important than having the right theory, until the final phase of the discussion, where, for example, the science demo is run and there is consensus on the outcome. (At that point, typically the teacher's role changes, and a focus on correctness, getting the right solution, and actively explaining to the students how to think about the situation takes place.) The teacher is primarily a coach, whose job is not to talk the students out of their home-based knowledge and the theories implicit in that knowledge, but rather to help them to explicate, clarify, and sharpen their theories. [15] (See O'Connor, 2001; Michaels & Sohmer, 2001a, 2001b; Michaels, O'Connor, & Resnick 2001 for more discussion of position-driven discussions.)

In the I-Club, by virtue of a set of carefully selected, collectively experienced theorizable situations (the "demos"), the problem space and students' theorizing is highly constrained; at the same time, these "position-driven discussions" are conducive to developing a variety of new discursive skills.

A range of turn-taking strategies promote equitable participation (such as round-robin tallies of all students' positions, student speakers' handing off "the floor" to someone of the opposite sex, etc.). Since everyone must commit to a position (even if one's confidence in one's theory is shaky and even if ultimately proven wrong), everyone is "in the game," and credited with being a theorizer and holder of scientific positions and arguments, well before one is an expert in the domain. This has the effect of promoting an uncharacteristic willingness among students (especially school-averse students) to participate actively and take risks in arguing for unpopular positions. Everyone is kept on the same page, but in the words of the I-Club teacher, "it's a page so big, no one can fall off." Keeping the page so big that no one can fall off is an interactional achievement, accomplished in part by the nature of the demos, and in part by skillful orchestration on the part of the teacher.

For each demo, the multiple outcomes can be imagined as the experimental set-up is poised, but not set into motion. The "game" is both to predict what will happen (or, post-hoc, to explain why some particular outcome occurred) and to muster your best, most persuasive, argument for your position. Each student, as he or she goes on line with a position or prediction, is in a sense running his or her own experiment. At the same time, the discussion allows students to hear others' reasoning and to build on others' ideas or to take from the group the most compelling evidence or ways of formulating an argument. This format thus promotes willing and universal participation among students (simply voting is a form of participation) and a great deal of discourse modeling and practicing of discipline specific ways of speaking.

PDDs bear a striking resemblance to central activities of constructivist mathematics pedagogies that center on group discussion of a single problem (e.g., Victoria Bill, 1993; Maggie Lampert & Deborah Ball, 1998; Cobb, P., Wood, T., & Yackel, F., 1993; O'Connor, 2001). Likewise, whole group discussion around a single rich "mathematizable situation" is common in Japanese math classes (as described by Stevenson & Stigler, 1992, and demonstrated in the TIMSS video tape of Japanese eighth grade math lessons in geometry and algebra).

In the domain of English Language Arts, The Great Books Foundation has developed a critical thinking curriculum that uses (what we call) position-driven discussion as the core pedagogical practice. The teacher's guide states that the most important element of a successful discussion is the framing question - defined as a question to which there are at least two reasonable and compelling answers.

While Michaels & Sohmer (2001) and O'Connor (2001) coined the term "position-driven discussion," they did not, of course, invent the genre itself. It has been described, discussed, and researched (to some degree) by others interested in what might be broadly thought of as collective sensemaking. Nicholas Burbules (1993) describes four different types of dialogue: dialogue as conversation, dialogue as inquiry, dialogue as debate, and dialogue as instruction. Position-driven discussions resemble Burbules' "debates": "The debating spirit of this form of dialogue is in its dedication to contrasting the merits of alternative positions from the strongest positions available. � [A] potential benefit can be in observing the respective cases made for each view and gaining a better sense of the number and strength of arguments available to them, pro and con on each position" (p. 119). Burbules locates the precursors to this form in the teaching of the rhetoricians in classical Greece. Seminal research on this form of talk as "collective comprehension activity" was carried out by Hatano and Inagaki (1991), based on an approach to teaching science in Japan developed by Kiyonobu Itakura, known as the "Hypothesis, Experiment, Instruction Method" or HEI. In this approach, a theorizable situation is set up with a framing question. Students vote, debate, and then revote.

Position-driven discussions are built into a number of different inquiry-oriented science curricula. In the US, these include Jim Minstrell's program in high school physics (1989); Eric Mazur's (1996) approach to college level physics, referred to as "the Peer Instruction Method." In an NSF funded project, "Assessing to Learn," the use of position-driven discussions prompted by a set of "items" serves as the heart of high school and college physics instruction. The program, like Mazur's, also incorporates the use of a "personal response system" (PRS), hand held clickers that allow students to go on line with their positions which are then transformed into a public histogram displaying student responses. This approach has been shown to promote high levels of student participation and productive group discussions in physics - discussions that simultaneously serve as opportunities for formative assessment by the teacher (Dufresne et al., 2000; Dufresne et al., 1996; Feldman, 1999). Each of these programs puts central emphasis on teacher-led, position-driven discussion with active student theorizing, debating, and voting for candidate positions.

How might position-driven discussions support learning science? First, position-driven discussions promote active participation before students are fully competent in a domain. In many cases, everyone is expected to commit to a position (and explain why) but it's perfectly fine to build on (or even copy) someone else's reasoning, provided the student states it in his or her own words. The teacher's role is to help students clarify and make explicit their position and the evidence for that position. Taken together, this kind of group discussion provides a great deal of support for students to listen to one another, build on one another's ideas, and take on new "ways with words," in the process. Because the goal is to have a well-developed position, which may or may not be right, students are willing to participate, even those who are most reluctant to participate in other kinds of group discussion. And each student is credited as a "player," a holder of a position or theory, whether or not his prediction or explanation wins out in the end. This provides a productive bridge from students' home worlds and everyday experiences to the more academic forms of reasoning, socialized in school.

There are, however, some potential difficulties with this kind of group discussion talk format. First, it works most easily with science demos or cases where the question on the table has a definitive answer, one that can be demonstrated or revealed to the students after their discussion and arguments. Ideally correctness rests in the world (in the science demo), rather than with the teacher. The teacher's role is to help students explicate their positions as clearly and cogently as possible (not indicating--even subtly--how close to the "right" answer they are). The teacher does NOT evaluate student contributions as right or wrong, as might be more common in other kinds of teacher-guided discussion or recitation. Rather, in position-driven discussions, the teacher typically scaffolds students by "revoicing" their contributions and pushing for clarification, so that everyone has access to everyone else's reasoning. Having a good "sayable" theory or position is more important than having the right theory, until of the end of the discussion, where, for example, the science demo is run and students see the actual outcome. At that point, the teacher's role changes, and a focus on correctness, getting the right theory, and actively explaining to the students how to think about the situation takes place.

Secondly, coming up with good framing questions is not easy. A PPD requires both a well-designed task and a carefully constructed framing question which will provoke a range of reasonable positions, no one of which is obviously correct. Additionally, the question must be carefully selected and sequenced among other tasks so as to advance the thinking of the group as a whole. It is unreasonable to expect a teacher to develop framing questions for PDDs without the support of a rigorous, coherent curriculum that emphasized student reasoning and inquiry through group discussion.

A third potential difficulty is that position-driven discussions crucially involve the teacher's active role in orchestrating, eliciting, and scaffolding students' predicting and theorizing. In order to pull off a "position-driven discussion" so that it engenders productive theorizing and learning, a teacher has to be quite knowledgeable in the content domain, and should know what kinds of assumptions learners are likely to make about the shared demonstration or problem. In guiding the discussion, the teacher must make productive use of students' non-technical, everyday experiences as well as their observations and experiences of the current problem. The teacher has to leverage agreements and disagreements about guesses and (later) hypotheses into intense and exciting classroom discussion. The demands on teacher knowledge and expertise in the domain are arguably lower in situations where teachers support students to develop their own questions (to which the teacher might well not know the answer) and to carry out a number of different investigations in small groups. In these cases, the teacher can model the stance of a learner, who asks good questions and guides students in the process of developing credible procedures for answering them. This is less the case in position-driven discussions. Although well-designed and well-sequenced science demos (e.g., I-Club demos discussed above) can carry a great deal of "intelligence" about the science under investigation, teachers must know the domain well in order to facilitate coherent and academically productive position-driven discussions.

Taken together, how do these three types of cultural tools interact to support the learning of science and the socialization of motivation? Position-Driven Discussions around appropriate demos have a built-in logic that makes certain kinds of intellectual work happen naturally.

Position-driven discussions -- by design -- require students to take a position or make a prediction. This automatically entails having some basis (a partial model or explanation), even before the student is aware that he or or she has one or precisely what it looks like. In this way, a position-driven discussion embodies and enacts key intellectual practices, developing and weighing scientific arguments.
Once you have a position (even before saying anything), you are fully in the game as a player. There is thus a low entry threshold for legitimate participation.
Being in the game, even if just raising a hand in a vote, positions you as a scientific "theorizer" (regardless of how cogent your argument is, or how successful the teacher is at interpreting and extending your contribution).

The demos recruit kids' everyday experience and ways with words. The kids are drawing on their embodied knowledge of the world, actively theorizing, using everything they can muster -- balloons, anchors, extracts from experience to help them explain their predictions. These are characteristics of science in action. Students find academically productive ways to be connected to one another. As these kids are being inducted into the I-Club practice through PDDs, they are learning to take each other seriously. They are following each other's arguments carefully, unpacking them, challenging premises and the appropriateness of explanatory analogies. They're learning that having a good argument is the point, not who said it first. They are learning to listen to one another and treat each other as "colleagues," not middle school peers.

Membership precedes competence.
You're in the game, you're a member before you're competent. The demos prompt students to use their everyday knowledge and experience. But because they are confronted with opposing views and then the outcome of the demo, they are prompted to reexamine and reintegrate their knowledge, bringing their everyday understandings into alignment with canonical scientific understandings. Although every student is trying to persuade others that they have the right answer by explicating their argument, the emphasis in the discussion is on having a well-explicated argument. Then the demo renders one position superior to the others.

A position-driven discussion doesn't rely on the teacher understanding every student contribution all of the time. The positions students take and the arguments they muster are often non-canonical and implicit. Both factors make their contributions hard for the teacher to understand on the fly. Fortunately, the activity is designed to support the iterative explication and expansion of positions, so misunderstanding is productive: providing rich opportunities for deeper digging, requests for clarification, evidence, and warrants.

In a position-driven discussion, the talk and task (structure and content) are inextricably linked.
Two final points characterize this linkage:

A large part of the collective competence achieved is carried in the talk/task amalgam of the position-driven discussion itself, not in the individual students or teacher.
Having a good "demo" is critical, but it is not enough. It has to be linked to explanatory tools and a talk format that mediates both students' and teacher's actions.

Conclusion

This is the beginning of a theorized account of how a practice promotes competence, connectedness, and autonomy (so that the Discourse comes to speak through the participants - students and teachers alike). If we are to disseminate ("go to scale") with the Investigators' Club practice or other "discourse intensive" pedagogies and programs, we need to understand better this level of discursive organization, the interplay of task structure, talk, and explanatory tools, and their role in socio-cognitive apprenticeship -- for both students and teachers.

References

Arons, A. (1997). Teaching introductory physics. New York: John Wiley and Sons.

Burbules, N. (1993). Dialogue in teaching: Theory and practice. New York: Teachers College Press.

Cobb, P., Wood, T., & Yackel, F. (1993). Discourse, mathematical thinking, and classroom practice. In N. Minick, E. Forman, & A. Stone (Eds.). Education and mind: Institutional, social, and developmental processes. Oxford, England: Oxford University Press.

Dufresne, R.J., Gerace, W.J., Mestre, J.P., & Leonard, W.J. (2000). ASK-IT/A2L: Assessing Student Knowledge with Instructional Technology. UMPERG technical report.

Dufresne, R.J., Gerace, W.J., Leonard, W.J., Mestre, J.P., & Wenk, L. (1996) Classtalk: A classroom communication system for active learning. Journal of Computing in Higher Education 7(2), 3-47.

Feldman, A. (1999). The role of conversation in collaborative action research. Educational action research, 7(1), 125-144.

Hatano, G. & Inagaki, K. (1991). Sharing cognition through collective comprehension activity. In L. Resnick, J. Levine, and S. Teasley (Eds.). Perspectives on socially shared cognition (pp. 331-348). Washington, DC: American Psychological Association.

Heisenberg, W. (1965). The Role of Modern Physics in the Development of Human Thinking. In F. T. Severin (Ed.), Humanistic viewpoints in psychology. New York: McGraw Hill.

Lampert, M., & Bell, D.L. (1998). Teaching, multimedia and mathematics: Investigations of real practice. New York: Teachers College Press.

Latour, B. (1986). Visualization and cognition: Thinking with eyes and hands. Knowledge and Society: Studies in the Sociology of Culture Past and Present. Vol. 6, pp. 1-40.

Latour, B. (1987). Science in Action. Cambridge, MA: Harvard University Press.

Lave, J. (1988a). Cognition in Practice. Cambridge: Cambridge University Press.

Lave, J. (1988b). The culture of acquisition and the practice of understanding. (Report No. IRL 88-0007). Palo Alto, Ca.: Institute for Research on Learning.

Lave, J. and Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

Leer, M., Bill, V., & Resnick, L. (1993, April). Mathematical power and responsibility for reasoning: Forms of discourse in elementary mathematics. Paper presented at the Annual Meeting of the American Educational Research Association, Atlanta, GA.

Liem, T.L. (1987). Invitations to science inquiry (2nd ed.) Chino Hills, CA: Science Inquiry Enterprises.

Mazur, E. (1996). Peer instruction: A user's manual. Newark, NJ: Prentice Hall.

Michaels, S. & R. Sohmer. (2001). "Discourses" that Promote New Academic Identities. In Li, D. (Ed), Discourses in Search of Members, pp. 171-219. New York: University Press of America.

Minstrell, J. (1989). Teaching science for understanding. In L.B. Resnick & L.E. Klopfer (Eds.), Toward the thinking curriculum: Current cognitive research pp. 131-149. Alexandria, VA: Association for Supervision and Curriculum Development.

O'Connor, M.C. (2001). "Can any fraction be turned into a decimal?" A case study of a mathematical groups discussion. Educational studies in mathematics, 46, pp. 143-185.

O'Connor, M.C. & Michaels, S. (1996). "Shifting participant frameworks: Orchestrating thinkng practices in group discussion." In D. Hicks (Ed.), Child Discourse and Social Learning. Cambridge: Cambridge University Press. Pp. 63-102.

Polanyi, M. (1962). Personal Knowledge: Towards a Post-Critical Philosophy. Chicago: University of Chicago Press.

Resnick, L. & Nelson-LeGall, S. (1999). Socializing intelligence. In, L. Smith, J. Dockrell, & P. Tomlinson (Eds.), Piaget, Vygotsky and beyond. London: Routledge.

Rogoff, B. (1990). Apprenticeship in Thinking: Cognitive Development in Social Context. New York: Oxford University Press.

Rommetveit, R. & Blakar, R.M. (Eds.) 1979. Studies of Language, Thought, and Verbal Communication. London: Academic Press.

Rommetveit, Ragnar. (1974). On Message Structure: A Framework for the Study of Language and Communication. New York: Wiley.

Rommetveit, Ragnar. (1992). Outlines of a dialogically based social-cognitive approach to human cognition and communication, in Astri Heen Wold (ed.), The Dialogical Alternative: Towards a Theory of Language and Mind. Oslo: Scandinavian University Press.

Schmidt, W., McKnight, C., & Raizen, S. (1997). A Splintered Vision: An Investigation of U.S. Science and Mathematics Education. Boston: Kluwer Academic Publishers.

Sohmer, R. (2000). "A Page So Big No One Can Fall Off": Apprenticeship as the Architecture of Intersubjectivity in an After-School Science Program for Inner City Middle School Students. (Unpublished Dissertation, Clark University).

Stevenson, H.W. & Stigler, J.W. (1992). The learning gap: Why our schools are failing and what we can learn from Japanese and Chinese education. New York: Simon & Schuster.

Third International Mathematics and Science Study (TIMSS). (1996). Boston College, TIMSS International Study Center. Chestnut Hill, MA: International Association for the Evaluation of Educational Achievement.