Date: Sat, 14 Oct 1995
From: Angel M.Y. Lin (mylin+@oise.on.ca)
Subject: Review of Griffin et al.'s formative experiment
(Chapter 5 of Contexts For Learning)

Hi fellow xmca-ers,

Attached below is my review of Griffin et al.'s chapter (chapter 5) in the book, Contexts For Learning. I started out to write a short review but alas ended up writing an extended essay review (7 pages single-spaced with references and footnotes). To provide you with an advanced organizer for reading this essay review, I'll outline the main points of my review as follows:

(1) Griffin et al.'s accomplishment: Discovering the computer program as a bona fide (albeit hidden) classroom co-participant, and illustrating ways of revising the program to make it a more liberal and flexible co-interactant... i.e., creating and reconstituting the possible constraints and resources that the computer program can impose on and afford to classroom interactions.

(2) Addressing the problem of institutionalizing innovative computer programs (or other artifacts and practices) in the school context:

(i) The need to consider the existing larger embedding socio-cultural, institutional, normative and motivational contexts for the incorporation of innovative artifacts/practices: taking a situated and activity theoretical perspective (Ref.: Engestrom)

(ii) The need to design "formative experiments" based on findings from "naturalistic observations" in a particular context,

(iii) The need for partnership and mutual communication between researchers and practitioners in the zone of proximal development of educational contexts: the mutual reconstitution of researchers and school practitioners.

In the following sections, I shall first discuss the importance of what the authors have achieved so far. Then I shall discuss the problem of institutionalizing computer programs in the school context, and propose some re- orientations in the conceptualization of, and research approaches to, contexts for learning. These re-orientations may hold better promise of helping us to understand what may facilitate and motivate the local (re)constitution of innovative contexts for learning in each unique socio- cultural situation.

COMPUTER PROGRAM AS A CO-PARTICIPANT IN CLASSROOM INTERACTIONS
The thrust of the authors' analysis is that a computer program in all its interactional consequences functions as a bona fide (albeit "hidden") co-participant in classroom interactions. As such, it can be a domineering co- participant who imposes pre-suppositions about "the definition of the situation" (i.e., "the context") and overly constrains other participants' actions, their understanding of the task situation, and subsequent interactions. Drawing on the theoretical and methodological resources of ordinary language philosophy and conversation analysis, the authors show how these consequences can happen, and how the researchers' collaborating with the software programmers to revise the program can change this computer-participant into a less domineering and more flexible interactant, e.g., by making available more menu and procedural options to teachers and students who then can choose from or re-sequence/design these options so as to make the computer-participant more responsive to the local needs of individuals and individual classroom situations (see examples discussed in pp. 136-143). The whole chapter is mainly about the discovery of those unfavorable interactional consequences of the behavior (so to speak) of the computer-participant and how the computer-participant has been gradually changed (by reprogramming) into a more liberal (so to speak), flexible co-participant who is more open to negotiations with other interactants (e.g., teachers, students).

This is a very important insight as far too often it is assumed that one single perfect version of a computer program can be designed for use for most school populations in most situations at most times. What the authors do in this chapter is to heighten the awareness of educational software programmers and researchers in the enterprise of designing and implementing computer-mediated learning and teaching in schools. Their message is forceful and multi- faceted: take a local, situated, interactive, multi-media perspective on computer-mediated learning and instruction. In practical terms, it means: developing and designing computer programs that are flexible and responsive to the needs of local, particular classroom situations, and programs whose interactional behavior does not constrain the subsequent interpretations and interactions of students in unfavorable directions (e.g., language that leads a student to interpret the re-doing of a task boring [pp. 140-141], a slow picture change that invites context boundary interpretations right in the middle of a task [pp. 141-142], a task sequencing that mispresupposes the wrong task [the two-step ponds rather than three-step ponds] to be the genetically primary example [p. 137]), and using the computer program in combination with auxiliary means of various modalities, e.g., drawing diagrams, representations, role-playing (pp. 143-144).

While the authors focus on foregrounding the possible constraints and resources that the hidden computer- participant can impose on and afford to classroom interactions1, and how the computer program can be revised to make it a more flexible and helpful co-participant, the authors have left out the larger socio-cultural, institutional and motivational contexts in which the classroom interactions are embedded. It seems that these are the areas where the Russian and American situations may be expected to differ considerably as the two countries may very well have very different educational and assessment systems, curricular organization and goals, computer technology cultures, school cultures, teacher cultures, student cultures, educational research cultures, etc.. A study of these larger embedding contexts and their relationships to the micro-level classroom interactions may seem too complex and broad to be handled in one single chapter. These larger embedding contexts, however, cannot be taken for granted if we are to ultimately address the practical question of how one goes about institutionalizing an innovative computer program for math learning in a particular school in a particular society and culture at a particular time.

INSTITUTIONALIZING COMPUTER-MEDIATED LEARNING:
SITUATED AND ACTIVITY THEORETICAL PERSPECTIVES ON CONTEXTS FOR LEARNING
One type of questions that one may ask the authors is: What is the local curricular context of the computer program, The Pond? How is the math curriculum organized in the schools studied, and how does The Pond fit in? What are the curricular goals in each school setting and in what ways can The Pond be seen (not only by the researchers or computer program designers but also) by the respective Russian and American school personnel (e.g., school principals, teachers, math subject panel chairpersons, parent-group representatives, etc.) to facilitate achieving their math curricular goals? etc..

A second type of questions that one may ask is: How will the students' math learning be assessed at school, what value will the school assessment system assign to what the students have learnt (implicitly and explicitly) with The Pond? What value will parents, employers, or the larger society assign to it? Do they assign differential values to males' and females' math achievement, or computer literacy? etc..

Seeking answers to these questions would lead one to pay attention to the normative and motivational context for incorporating a computer program into the math curriculum, the classroom and the school. The research question then becomes: Is there a normative and motivational context for the institutionalization of math learning mediated by The Pond in this particular school in this particular society and culture at this time?

However, it seems that to ask the above research question is still to put the cart before the horse, for why on earth would any class, any teacher, any school in any society and culture be interested in incorporating The Pond, or The Lake2, or any other computer program in the first place? Why should they be convinced of its value and usefulness in their own local situations in the first place? One cannot always expect any particular school to be motivated to experiment with whatever latest educational computer software available because some expert educational researchers or software developers think that this computer program may be a great way of mediating math learning and teaching or because it has proved to work in some other schools. The researchers' focus then should not be on The Pond, or any given computer program, but instead on the particular school context that the researchers are interested in working in. Their research questions should then become: What are these people doing and how are they doing it? What difficulties or limitations are they experiencing? What kind of new artifacts (including computer programs) or practices will best improve their current learning and teaching situation with reference to THEIR OWN curricular goals and socio-cultural values? What alternative goals or values can we as researchers communicate to them so as to provide them with alternative options to consider?

The above issues, however, cannot be addressed without FIRST studying and understanding the existing math- learning/teaching/assessment practices, and curricular and administrative organization patterns in any particular school, as well as the socio-historical and cultural values of the society in which the school is situated. Daunting though such a research task seems to be, Engestrom's development of activity theory (1987, 1990a, 1990b, 1993) provides a theoretical framework and a research methodology that may be useful in our situation. From this activity theoretical perspective, new artifacts, tools, and innovative ways of working cannot be introduced without a recognition and understanding by the participants themselves (in JOINT exploration with the researchers) of how the new artifacts or innovative ways of working can resolve EXISTING inner contradictions in their present activity system. Their recognition of, and desire to find solutions to, existing problems or limitations serve as their motivation for developing, accepting, adapting and experimenting with new artifacts and innovative approaches in collaboration with the researchers. The distance between their present activity system and the new expanded activity system where new artifacts, tools and practices are experimented with is the zone of proximal development for their educational context. This zone of proximal development, however, cannot be defined or imposed from without. Nor can the new artifacts be introduced solely from without. They have to be discovered, developed, accepted or adapted by those participants in the particular contexts in JOINT research and exploration with the researchers. In other words, researchers and computer programmers cannot simply transmit what they consider useful to local school practitioners, just as teachers cannot simply transmit what they consider to be useful knowledge to their students. Researchers should be engaged in a JOINT discovery and learning process with school practitioners.

NO "FORMATIVE EXPERIMENTS" VERSUS "NATURALISTIC OBSERVATIONS"
The implication of the above perspectives is that "formative experiments" and "naturalistic observations" (as defined by the authors in pp. 123-125) are not two largely unrelated research approaches to educational reform, as the authors seem to have implied (p. 124). The authors rightly point out the limitations of naturalistic observations alone (p. 124). However, one also has to recognize that the design of any formative experiment must be informed by the findings of naturalistic observations of the existing local particular situation in which the formative experiment is to be launched.

For instance, the authors mention Mehan et al. (1983, p. 226)'s comment, "Why should we expect that the same practices that have produced wide-spread academic failure will create propitious environments for computer use?", as the reason for the need to "go beyond observation and create contexts to study computers in education" (p. 124). The comment I believe can be re-phrased the other way round: "Why should we expect The Pond, or any particular computer program, to be able to help change those existing practices that have produced wide-spread academic failure?". In other words: What are the existing practices in this particular context? How does the introduction of a particular computer-participant change those unfavorable practices? How can we design this computer-participant to specifically help change those practices?

To illustrate, let us consider this hypothetical example. If it has been discovered in a particular classroom (by some naturalistic observations) that girls and boys engage in a particular type of division of labor when they collaborate on solving math problems, e.g., typically boys do the reasoning and problem-solving part while girls do the work of copying and tidying up the solutions and the oral presentation of the solutions, then this finding should inform the design of the subsequent formative experiment to be launched in this particular situation, e.g., designing a computer program to be introduced into this classroom such that this computer-participant will constrain the division of labor by boys and girls in a manner different from the existing pattern uncovered by naturalistic observations.

CONCLUSION:
THE MUTUAL RECONSTITUTION OF RESEARCHERS AND SCHOOL PRACTITIONERS
However, even when researchers have designed formative experiments based on findings from naturalistic observations, there can be little success of institutionalizing the innovative artifacts and practices if all these are conducted and introduced from without by the researchers alone. As mentioned earlier, researchers cannot simply "transmit" what they consider useful to local practitioners. Partnership, joint exploration and mutual communication seem to be the key elements in their working together in the zone of proximal development of educational contexts. Since contexts are always constituted locally by the local participants, there can be little reconstituting of educational contexts without the active reconstituting of the local participants themselves. Such reconstitution (e.g., of one's values and practices), however, can take place only as researchers and school practitioners are engaged in mutual reconstitution in a process of mutual communication and joint exploration. While Griffin et al. may not have addressed all these issues in their chapter, their detailed examples of how teachers, researchers and computer programmers collaborate to revise The Pond based on their local observations of its interactional consequences for the classroom participants help provide a micro-example of what also can and needs to be done in the larger embedding contexts for the successful institutionalization of any innovative artifacts or practices, including a computer program.

NOTES
1. I have deliberately changed the authors' phrase "creating and reconstituting contexts for educational interactions" to "creating and reconstituting constraints on and resources for classroom interactions" in this sentence. The reason is that context is a local accomplishment achieved in each unique situation through the dynamic negotiation work of the interactants in that situation. What the authors describe in their chapter are the many different constraints and resources that the computer- as-participant can impose on and afford to the other human interactants in their local co-construction of the context, and how these constraints and resources can be changed beforehand (by reprogramming) to facilitate the participants' co-construction of the context in educationally desirable directions. Likewise, my substitution of "classroom interactions" for "educational interactions" is motivated by the observation that not all interactions that can take place among the children, teacher and computer are necessarily "educational". "Educational interactions" are again a local accomplishment by the co-participants in each unique situation and cannot be guaranteed in advance (though may be facilitated) by the availability of the best kinds of artifacts and resources (e.g., well-designed computer program and other auxiliary means).

2. There is no such a computer program called The Lake. I made this up just to emphasize my point.

REFERENCES
Engestrom, Y. (1987). Learning by expanding: An activity- theoretical approach to developmental research. Helsinki: Orienta-Konsultit.

Engestrom, Y. (1990a). Activity theory and individual and social transformation. Opening address at the Second International Congress for Research on Activity Theory, Lahti, Finland, May 21-25, 1990.

Engestrom, Y. (1990b). Learning, working and imagining: Twelve studies in activity theory. Helsinki: Orienta- Konsultit.

Engestrom, Y. (1993). Developmental studies of work as a testbench of activity theory: The case of primary care medical practice. In Chaiklin, S. & Lave, J. (Eds.) Understanding practice: Perspectives on activity and context, pp. 64-103. Cambridge: Cambridge University Press.