Computer-Supported Collaborative Learning: A Script Theory of Guidance

Abstract

This article introduces a script theory of guidance designed for Computer-supported Collaborative Learning (CSCL). Comprising four types of components—play, scene, role, and scriptlet—for both internal and external scripts, alongside seven core principles, this theory explores how CSCL practices are shaped by the dynamically adjusted internal collaboration scripts of learners. Furthermore, it elucidates the development of these internal collaboration scripts through active participation in CSCL practices. The theory highlights the critical role of actively applying subject matter knowledge within CSCL contexts and prioritizes transactive knowledge application to enhance learning. It also explains how external collaboration scripts influence CSCL practices and the evolution of internal collaboration scripts. The principles delineate an optimal scaffolding level for external collaboration scripts, enabling the formulation of hypotheses regarding their gradual removal. Finally, the article identifies key conceptual challenges and areas for future research in computer-supported collaborative learning.

Collaborative learning is employed as an instructional method for numerous reasons. From a cognitive standpoint King, 1997, activities inherent in collaborative learning, such as explanation and questioning, are believed to modify participants’ cognitive structures. Sociocultural perspectives suggest that through collaborative engagement, individuals gradually internalize collaborative practices, developing collaboration skills and cognitive strategies applicable in diverse contexts Kolodner, 2007. However, research consistently indicates that learners often do not naturally engage in these advanced collaborative processes without structured guidance Weinberger, Stegmann, Fischer, & Mandl, 2007. Consequently, a central research question is how to effectively support collaborative learning to foster high-level collaboration and improved learning outcomes. Computer-supported collaborative learning (CSCL) introduces novel educational opportunities that many students have not previously experienced. Examples include web-based inquiry learning in classrooms Linn & Slotta, 2000, online case discussions Weinberger, Ertl, Fischer, & Mandl, 2005, and collaborative wiki writing and editing Cress & Kimmerle, 2008. However, the more CSCL practices diverge from traditional educational experiences, the greater the challenge for students to collaborate effectively. Learners lacking prior experience with these collaborative methods may not possess the necessary knowledge to guide their collaboration in these environments. Following Kollar, Fischer, and Slotta 2007, we term this knowledge “internal collaboration scripts.” These scripts are pivotal in the script theory of guidance in CSCL, providing insights into what directs learner actions in CSCL settings.

External collaboration scripts King, 2007; Kollar, Fischer, & Hesse, 2006; Rummel & Spada, 2005; Weinberger et al., 2005 offer a way to address the absence or inadequacy of internal collaboration scripts by providing collaborators with guidance in collaborative situations. Despite descriptive analyses of CSCL scripts Kobbe et al., 2007 and substantial empirical research on the effects of external collaboration scripts on learning processes and outcomes Hämäläinen, Oksanen, & Häkkinen, 2008; Rummel & Spada, 2005; Schellens, van Keer, De Wever, & Valcke, 2007; Stegmann, Weinberger, & Fischer, 2007, a unified theoretical framework—a script theory—remains underdeveloped. This article aims to outline a script theory of guidance in CSCL to fill this theoretical gap. The theory posits that internal and external collaboration scripts are central to guiding learner activities in CSCL environments. These scripts are considered distinct yet structurally parallel. An internal collaboration script is a dynamically configured knowledge structure about a collaborative practice and its components at varying complexity levels (internal collaboration script components). It guides an individual’s understanding and actions within a collaboration, evolving dynamically from its elements during a specific collaborative instance. Conversely, an external collaboration script is a set of representations (textual or graphical) of a collaborative practice and its components, also at potentially different complexity levels (external collaboration script components). Delivered by an external source, such as a teacher or a website interface, it aims to guide learners’ collaborative activities. This guidance arises from how these external representations influence collaborators’ configurations of internal collaboration script components.

The script theory of guidance is structured around four script components and seven principles. These seven principles, detailed in Table 1, are categorized to answer three core questions:

• 1. How do internal collaboration scripts and situational characteristics interact to shape CSCL practices (Principles 1 & 2)?
• 2. How do internal collaboration scripts evolve and adapt through CSCL participation, and what role do they play in knowledge acquisition (Principles 3–5)?
• 3. How do external collaboration scripts impact CSCL practices and individual knowledge acquisition (Principles 6 & 7)?

TABLE 1.

Principles of the Script Theory of Guidance in Computer-Supported Collaborative Learning

Principle No.	Principle Name	Principle Formulation
1	Internal script guidance principle	When participating in a CSCL practice, the learner’s understanding of and acting in this situation is guided by dynamically configured and reconfigured internal collaboration scripts consisting of play, scene, scriptlet, and role components.
2	Internal script configuration principle	How an internal collaboration script is dynamically configured by a learner from the available components to guide the processing of a given situation, is influenced by the learner’s set of goals and by perceived situational characteristics.
3	Internal script induction principle	If a learner participates in an initially unfamiliar CSCL practice, then he or she builds a new configuration of already available internal script components and, through repeated application of this configuration of internal script components, develops new higher-level components (play, scene, or role) that organize the subordinate components (scenes, roles, and scriptlets) for this CSCL practice.
4	Internal script reconfiguration principle	If a learner’s employed internal collaboration script (i.e., a configuration of internal script components) does not lead to understanding or successful actions in a CSCL practice, the internal collaboration script configuration is likely to be modified.
5	Transactivity principle	The more a given CSCL practice requires the transactive application of knowledge, the better this knowledge is learned through participation in this CSCL practice.
6	External script guidance principle	External collaboration scripts enable learners to engage in an instance of a CSCL practice at a level beyond what they would be able to without an external collaboration script either by inhibiting the automated use of internal script components or by inducing the application of internal script components that are not yet organized by a specific higher level script component.
7	Optimal external scripting level principle	An external collaboration script is most effective for knowledge acquisition if it is directed at the highest possible hierarchical level of internal collaboration script components for which subordinate components are already available to the learner.

Open in a new tab

Note. CSCL = computer-supported collaborative learning.

These four components—play, scene, role, and scriptlet—are fundamental and are introduced alongside the initial set of principles.

The script theory of guidance in CSCL is rooted in two primary theoretical frameworks. Firstly, it draws upon schema-theoretical accounts of cognition to explain the development and application of internal scripts in understanding and acting collaboratively. The script concept has evolved significantly in cognitive science, from early notions of large, stable cognitive structures Schank & Abelson, 1977 to contemporary approaches emphasizing dynamic memory components that are flexibly combined based on situational aspects and individual goals Kintsch, 1998; Schank, 1999. This theory builds on dynamic memory theory Schank, 1999, a more adaptable schema theory. Secondly, the script theory integrates sociocultural perspectives, adopting the view that social discourse precedes and shapes complex cognitive skills—the Genetic Law of Development Vygotsky, 1978. It also incorporates the Zone of Proximal Development Vygotsky, 1978, suggesting individuals advance knowledge and skills through participation in activities slightly beyond their current level, supported by more knowledgeable peers. In this theory, sociocultural ideas primarily explain how CSCL participation stimulates, but does not dictate, individual cognitive development Kolodner, 2007. A key critique of sociocultural approaches has been the lack of detail regarding the structure and selection of internalized knowledge in subsequent discourse. By merging sociocultural accounts with schema theory assumptions, the script theory of guidance addresses these criticisms. Another criticism of sociocultural approaches is the underspecified nature of internalization. To this end, the theory specifies (a) how internal collaboration scripts are induced and reconfigured through CSCL participation and (b) how external collaboration scripts can be designed and gradually removed to facilitate internal collaboration script development.

Empirical findings are presented throughout the theory’s principles to provide illustrations and preliminary support. While many cited studies originally offered alternative theoretical interpretations, the script theory of guidance offers a cohesive perspective to explain a wide array of findings in CSCL research. However, the presented studies should not be considered a definitive empirical validation of the script theory of guidance’s assumptions.

HOW INTERNAL COLLABORATION SCRIPTS SHAPE CSCL PRACTICES

The script theory of guidance in CSCL includes two principles detailing how internal collaboration scripts influence CSCL practices and their selection (Table 1):

• 1. When participating in a CSCL practice, the learner’s understanding and actions are guided by dynamically configured and reconfigured internal collaboration scripts, composed of play, scene, scriptlet, and role components (Principle 1, internal script guidance principle).
• 2. The dynamic configuration of an internal collaboration script from available components, guiding situation processing, is influenced by the learner’s goals and perceived situational characteristics (Principle 2, internal script configuration principle).

CSCL practices are defined as recurring verbal and nonverbal interaction patterns among two or more participants in technology-enhanced settings with similar functions. Knowledge of such practices constitutes an “internal collaboration script,” viewed as configurations of knowledge components about a collaborative practice. The term “script” intentionally evokes theater, as highlighted by Schank 1999. However, our theory extends dynamic memory theory Schank, 1999 by more explicitly integrating and defining a role component and by replacing Schank’s “memory organization package” (MOP) with a play component to emphasize the partial parallelism between internal and external collaboration scripts. Knowledge about a collaborative practice, therefore, comprises these components: (a) The play component represents knowledge of the overall “play” participants are engaged in (e.g., argumentative dialogue or joint Wiki writing), including the sequence of scenes and roles involved. (b) Scene components are knowledge about specific situations within a play Schank, 1999, p. 123. For instance, in an argumentative dialogue, a scene might be developing a counter-position to a previous claim. (c) Scriptlet components are knowledge of activity sequences within scenes Schank, 1999, p. 125. For example, in developing a counter-argument, a scriptlet might suggest first stating a claim and then providing supporting evidence. (d) Role components represent knowledge of roles within the collaborative practice, organizing appropriate activities for specific participants. Roles typically span multiple scenes and activities, intersecting with scene components in organizing scriptlets for individual activities. In an argumentative dialogue, the role component includes knowledge of at least two differing positions advocated by different individuals acting compatibly by exchanging arguments across scenes.

Internal collaboration scripts are not rigid cognitive structures directing a complete collaborative practice. Instead, they are flexible configurations of knowledge components (plays, scenes, scriptlets, and roles) with a likely sequence, adaptable dynamically to situational or activity changes. This process is often subconscious Schank, 1999.

According to Principle 1, a CSCL practice instance emerges from the interaction of participants guided by their individual knowledge, embodied in their internal collaboration script component configurations. Principle 2 states that the configuration and reconfiguration of internal collaboration script components depend on individual goals and perceived situational constraints and affordances, which evolve with the collaborative practice. Guidance from script component configurations influences not just actions but also situation understanding and expectations.

Evidence outside CSCL for the importance of internal scripts in understanding and acting comes from developmental psychology Lampinen, Faries, Neuschatz, & Toglia, 2000 and psycholinguistics Kellermann, Broetzmann, Lim, & Kitao, 1989. Kellermann et al. 1989 identified a culturally shared “initial conversation MOP” (p. 27) with scenes (initiation, maintenance, termination) and scriptlets (“subroutines,” p. 50), guiding students in first-time conversations.

Similarly, internal collaboration scripts guide learners in CSCL practices. Empirical support comes from a study examining how internal collaboration scripts affect collaborative argumentation quality in CSCL inquiry Kollar et al., 2007. Ninety secondary students participated, and results showed that students with more developed internal argumentation scripts (assessed by discourse protocol analysis) contributed more relevant arguments than those with less developed scripts.

Principle 2 emphasizes that situational affordances and constraints influence internal collaboration script component configuration. Affordances are perceived activity possibilities Norman, 1993. An empty text field with a cursor suggests text entry. Constraints are perceived activity limitations. An online forum that pre-fills “clarification question” in subject lines and limits reply box size may discourage elaborate counter-arguments Kirschner, Beers, Boshuizen, & Gijselaers, 2007.

Situational constraints and affordances vary with external representations, and specific representations can trigger corresponding internal collaboration script components. Suthers and Hundhausen 2003 found that university student dyads using matrix formats to represent data, hypotheses, and evidence in science explorations focused more on evidential relations in verbal exchanges than those using graph or text formats. This suggests that affordances for specific problem aspects influenced participants’ internal collaboration script component configurations.

Beyond situation, personal goals also shape internal collaboration script selection. Configurations aligning with current goals are more likely to be used. Pfister and Oehl 2009 studied 118 university students collaborating in chat groups, with either individual or group reward focuses. Group-focused students were better at grounding discussions using chat features to reference text. A group goal focus may lead to different internal collaboration script configurations than an individual focus, evident in improved grounding processes in the group goal condition.

DEVELOPMENT OF INTERNAL COLLABORATION SCRIPTS AND THEIR ROLE IN KNOWLEDGE ACQUISITION

Three additional principles of the script theory of guidance address the internalization of collaboration scripts and their role in knowledge acquisition (Table 1):

• 3. If a learner engages in an initially unfamiliar CSCL practice, they construct a new configuration of existing internal script components. Through repeated application, they develop new higher-level components (play, scene, role) that organize subordinate components (scenes, roles, scriptlets) for this CSCL practice (Principle 3, internal script induction principle).
• 4. If a learner’s internal collaboration script (component configuration) fails to produce understanding or successful actions in a CSCL practice, the configuration is likely to be modified (Principle 4, internal script reconfiguration principle).
• 5. The more a CSCL practice demands transactive knowledge application, the more effectively knowledge is learned through participation (Principle 5, transactivity principle).

Consider an individual encountering a novel collaborative situation. Initial confusion arises from lacking a functional play component to guide understanding and action. However, the individual possesses internal collaboration script components from other practices that share aspects (e.g., scenes). They will utilize these components (scenes, roles, scriptlets) from other practices that fulfill comparable functions. According to the internal script induction principle, repeated use of a new configuration of these components in similar instances of the unfamiliar CSCL practice leads to the development of a new play component, integrating previously unconnected, successfully employed subordinate components (Principle 3). The same process applies to forming new scene or role components.

However, applying components from other internal scripts is just one approach to unfamiliar situations. If the selected internal collaboration script components lead to unsatisfactory understanding or actions in a CSCL practice, an “expectation failure” Schank, 1999, p. 41 occurs, increasing the likelihood of reconfiguration (Principle 4).

Empirical support for the internal script induction principle (Principle 3) comes from Rummel and Spada 2005, where participants (N=72) engaged in an unfamiliar CSCL practice—interdisciplinary patient case work between psychology and medicine students via videoconference. Observing a model collaboration beforehand helped learners acquire knowledge about collaborating in this new CSCL practice. Observing a role model likely enabled learners to select previously unconnected internal collaboration script components, facilitating effective engagement in the unfamiliar CSCL practice. Repeated participation led to a new play component integrating these components.

Kapur and Kinzer’s 2009 study on productive failure supports the internal script reconfiguration principle (Principle 4). Science students (N=177) learned collaboratively with well- or ill-structured problems. After collaboration, they individually solved both problem types. Learners failed more often with ill-structured problems during collaboration, but those who failed outperformed those who did not on individual ill-structured post-tests. The authors argued that failure during collaboration helps discern relevant features and structures. In terms of script reconfiguration, learners experiencing expectation failures with ill-structured problems were more likely to reconfigure their internal scripts, improving subsequent problem-solving.

Environmental feedback is crucial for learning. However, learning is not automatic upon encountering failure Dole & Sinatra, 1998. Conditions for failure-induced script reconfiguration are not fully understood, but a promising hypothesis is that failures are most effective when hindering personally relevant goals Schank, 1999.

The transactivity principle (Principle 5) posits that greater transactive knowledge application in CSCL practices leads to better knowledge acquisition. This is based on two assumptions: First, active knowledge application enhances acquisition. Schank 1999 argued that without meaningful practices, subject matter knowledge may become encapsulated in scripts for exam reproduction, remaining inert and inapplicable outside exams Renkl, Mandl, & Gruber, 1996. Stegmann, Weinberger, and Fischer 2011 found that more subject matter knowledge applied in argumentative activities correlated with greater individual knowledge acquisition in 108 educational science students in online discussions.

Second, learning increases with the transactivity demanded by CSCL practices. Transactivity occurs when learners build on partners’ contributions, using them as resources. Dillenbourg and Jermann 2007 emphasized mutual reference as crucial for collaborative learning effectiveness (“Split Where Interaction Should Happen,” p. 292). Teasley 1997 defined transactive contributions as those developing another’s thought through integration, critique, clarification, or paraphrase. In a study (N=24), Teasley found a positive correlation between transacts and learning outcomes in joint problem-solving. Chi’s 2009 review showed better learning outcomes when partners used each other as resources and built on thoughts. Stegmann, Weinberger, et al. 2011 also showed that higher discussion transactivity resulted in more subject matter knowledge application and acquisition, supporting this principle.

EFFECTS OF EXTERNAL COLLABORATION SCRIPTS ON CSCL PRACTICES AND LEARNING

External collaboration scripts are representations that guide CSCL practices by either facilitating or inhibiting the application of internal collaboration script components. Facilitation occurs when external scripts create affordances influencing learners’ selection of specific internal script components (Principle 2), applied in CSCL practices (Principle 1). Inhibition reduces the likelihood of dysfunctional internal script components by (a) interrupting automated sequences through situational constraints and prompts, or (b) creating incompatible affordances that engage learners in sequences incompatible with initially selected internal scripts.

External collaboration script components largely mirror the four internal script component types. However, internal script component configurations are not mere reflections of external scripts. External scripts act as scaffolds inducing functional internal script configurations. The same scaffolds can stimulate different internal script configurations in learners at varying expertise levels. The theory distinguishes four scaffold types (external script components) targeting different cognitive levels: (a) Play scaffolds are prompts influencing the top-level configuration, defining general collaboration goals like joint problem-solving or argumentation without detailed guidance. (b) Scene scaffolds target known scenes (e.g., individual analysis, group discussion, presentation), sequencing them into a play. (c) Role scaffolds target role-based knowledge application, assigning specific roles. (d) Scriptlet scaffolds prompt application of scriptlets needed within scenes.

The third principle set addresses external collaboration script effects on CSCL practices and knowledge acquisition (Table 1):

• 6. External collaboration scripts enable learners to participate in CSCL practices at a level beyond their unscripted capability, by inhibiting automated internal script component use or inducing application of components not yet organized by a higher-level script component (Principle 6, external script guidance principle).
• 7. External collaboration scripts are most effective for knowledge acquisition when directed at the highest hierarchical level of internal collaboration script components for which subordinate components are already available (Principle 7, optimal external scripting level principle).

Numerous studies support the external script guidance principle (Principle 6) Hämäläinen et al., 2008; Kollar et al., 2007; Schellens et al., 2007; Schoonenboom, 2008; Stegmann et al., 2007; Wecker, Stegmann, et al., 2010; Weinberger et al., 2005; Weinberger, Stegmann, & Fischer, 2010. These studies show that external collaboration scripts improve CSCL discourse compared to unstructured CSCL. Schoonenboom 2008 found that scene scaffolds supporting grounding in CSCL (individual input, discussion, consensus) in distance learning groups (N=42) led to more contributions and better knowledge sharing and difference identification compared to unscripted learning. Hämäläinen et al. 2008 found similar positive effects of scene scaffolds in vocational student studies.

Schellens et al. 2007 used role scaffolds in asynchronous discussion boards, leading to more role-congruent activities. Scripted groups in two student cohorts (N=223 and N=286) showed higher knowledge-building activities in discussions than unscripted groups.

Stegmann, Wecker, Weinberger, and Fischer 2012 used scriptlet scaffolds in online discussions to support argument construction with claim, ground, and qualification. A graphical interface with text fields labeled “claim,” “ground,” and “qualification” was used. Scriptlet-scaffolded learners (N=48 in groups of three) showed greater cognitive elaboration and argumentation knowledge acquisition than unsupported learners.

The optimal external scripting level principle (Principle 7) suggests that scripts scaffolding already available subordinate internal script components are less efficient than those targeting higher-level components, due to unnecessary information processing. “Overscripting” Dillenbourg, 2002 relates to this. Overscripting is not just excessive scaffolding but occurs when external scripts hinder self-regulated application of appropriate higher-level internal script components—scripting at the wrong hierarchical level. Thus, unnecessary lower-level scaffolds hinder knowledge acquisition.

Mäkitalo et al. 2005 support this. Scriptlet scaffolds in asynchronous CSCL discussions for educational science students (N=48 in groups of three) impeded individual knowledge acquisition. Applying the optimal scripting principle, students may have already possessed adequate scriptlets, making scaffolds unnecessary and potentially causing cognitive overload.

Scripting at a level too low (too specific), as in Mäkitalo et al. 2005, can hinder subject matter knowledge acquisition and internal script internalization. Such scripts remove learner self-regulation, preventing higher-level internal script component development. Learners recently supported by external scripts are guided by induced internal script components (Principle 1), developing corresponding higher-level components (Principle 3). Optimal scripting levels shift higher with repeated application. Thus, Principle 7 implies learning is more likely when learners apply newly developed higher-level script components for self-regulation. Fading external collaboration scripts offers this opportunity.

Wecker and Fischer 2011 demonstrated fading’s influence. Learners supported by unfaded or faded scriptlet scaffolds for counter-argument writing in asynchronous discussions. Faded-script learners acquired higher quality internal argumentation scripts than unfaded, provided partners monitored strategy adherence. This indicates that fading, with continuous script application, can improve script internalization.

Continued adherence to the script-suggested strategy is crucial for fading to positively affect internal script development. Stegmann, Mu, Gehlen-Baum, and Fischer 2011 showed no positive effect from simply fading scripts without additional instructional means to ensure continued strategy use. Wecker, Kollar, Fischer, and Prechtl 2010 compared faded versus continuously available scripts for collaborative information search in genetic engineering learning. Fading was based on search count, not search quality, over weeks. Faded-script learners did not internalize information search scripts better than continuously scripted learners.

The optimal scripting level principle also implies that adaptive fading, tailored to learner development, is more effective than fixed fading. Tsovaltzi et al. 2010 adaptively adjusted scaffolds based on discussion quality. Adaptive fading outperformed non-adaptive and unscripted conditions in qualitative and descriptive quantitative results.

CONCLUSIONS AND FUTURE RESEARCH

This article outlined a script theory of guidance in CSCL. Internal collaboration scripts develop as learners attempt to understand and act effectively in recurring CSCL practices, which are shaped by these internal scripts. Flexible internal scripts dynamically adjust if they fail to produce understanding or action aligned with learner goals or situational changes. New internal script components emerge from successful reconfigurations of existing components. Transactive knowledge application in CSCL practices enhances knowledge acquisition.

External collaboration scripts, with play, scene, role, and scriptlet scaffolds, stimulate or inhibit internal script components, aiding learners in acquiring higher-level internal components and subject matter knowledge. Optimal external scripts encourage self-directed component use, targeting the highest possible level of internal components for which subordinates are available, avoiding redundant activity prescriptions. This theory informs the design of external collaboration scripts, indicating when they are effective, which scaffolds induce specific components, and how to remove scaffolds to promote self-directed script configuration and application.

CSCL studies show well-designed external scripts enable complex CSCL practices, fostering more elaborate internal collaboration scripts. Well-developed internal scripts are self-regulation tools for knowledge acquisition via CSCL. An open question is the relationship between subject matter knowledge and collaboration scripts. The theory suggests subject matter knowledge is better acquired through transactive CSCL use, supported by some findings Stegmann, Weinberger, et al., 2011. However, script induction/reconfiguration and subject matter knowledge acquisition may not always be synergistic. Pre-existing well-developed internal scripts benefit subject matter knowledge acquisition Kollar et al., 2007. Developing internal scripts before subject matter learning may be more effective O’Donnell & Dansereau, 1992, but this might underutilize CSCL’s real-time support. Some studies show process-related scripting can promote both subject matter knowledge and internal scripts simultaneously Wecker, Kollar, et al., 2010. Future research should directly investigate the relationship between script induction/reconfiguration and subject matter knowledge acquisition, varying internal script elaboration and analyzing subject matter knowledge association with internal scripts.

Another open question is the helpfulness of the current script components and levels (play, scene, scriptlet, role) for analyzing and designing all collaboration scripts. For example, joint problem-solving CSCL includes exploration, solution, and evaluation scenes Slof, Erkens, Kirschner, Jaspers, & Janssen, 2010. The current theory might conceptualize exploration as a scene with nested scenes. Exploring additional levels like an “act” component could further refine script analysis and design for specific CSCL practices.

The issue of overly high external CSCL support is also important. A misconception is that CSCL scripts are rigid interaction templates ensuring knowledge acquisition at the expense of natural collaboration. Collaboration scripts should be seen as temporary aids when higher-level internal scripts are lacking or not spontaneously transferable. They primarily prevent underscripting effects—limited learning due to overly open environments Kirschner, Sweller, & Clark, 2006. However, learners also learn from failure Kapur & Kinzer, 2009; Schank, 1999 when internal scripts are ineffective, as reflected in the reconfiguration principle. Designing learning environments to systematically enable productive failure in collaborative learning without withholding necessary guidance is an interesting challenge.

Fading is another crucial research area Wecker & Fischer, 2011. Adaptive and adaptable CSCL scripts are receiving significant research attention Diziol, Walker, Rummel, & Koedinger, 2010; Tsovaltzi et al., 2010. Adaptive fading could be researched from a developmental perspective. While required external script support may decrease for specific skills, collaborative practices may increase in complexity as children develop. Overall external scripting levels might remain constant over time, but the specific script content should adapt.

Future adaptive external collaboration script research and implementation require better, more efficient methods for diagnosing internal collaboration scripts. Schema and script research methods like recall and reaction times could be adapted. However, dynamic and adaptive nature of internal scripts must be considered. Computational linguistics advancements could diagnose dynamic internal scripts during CSCL use Mu, Stegmann, Mayfield, Rosé, & Fischer, 2012; Rosé et al., 2008. Machine-readability is another prerequisite for adaptive scripts. Script formalization, a challenging interdisciplinary endeavor, aims for psychologically valid and machine-readable descriptions of script components Hernandez-Leo, Jorrin-Abellan, Villasclaras-Fernandez, Asensio-Perez, 2010; Ronen, Kohen-Vacs, & Raz-Fogel, 2006.

Self-regulation and metacognitive awareness of internal scripts are also important. Increased internal script components potentially increase learner freedom, offering more understanding and action options. However, the theory does not explicitly address learner awareness of their internal scripts or metacognitive components enabling conscious script component selection. Relating script theory to shared and co-regulation research Järvelä & Hadwin, 2013 is a promising direction, with adaptable external scripts Wang, Kollar, Stegmann, & Fischer, 2011, where learners discuss script component selection, serving as a bridge.

On a broader theoretical level, this article proposes an instructional theory building on dynamic memory theory Schank, 1999 to address the rigidity of early schema theories in explaining adaptive social behavior. The theory extends dynamic memory theory by explicitly linking cognition to collaborative practice, both origin and application field. It draws on sociocultural ideas, particularly the Genetic Law of Development and Zone of Proximal Development Vygotsky, 1978. The script theory of guidance in CSCL has the potential to bridge cognitive theorizing in psychology with organizational and institutional collaboration research in social sciences. CSCL is a promising field to explore and strengthen this link.

ACKNOWLEDGMENTS

The authors contributed equally to this article and are therefore listed in alphabetical order. We thank Nicolas Balacheff, Jim Slotta, and Jan-Willem Strijbos for their critical and encouraging comments on earlier versions of the paper. We are grateful to Armin Weinberger and Heinz Mandl for more than a decade of inspiring collaboration that laid the foundation of this work.