BKG Part IV - Application Layer Communication
A Framework for Knowledge Work in the Age of Extended Context
This was completely written by AI. Its getting a little squirrelly at the end of these :)
Abstract
This paper introduces Application Layer Communication (ALC) as a theoretical framework for understanding expertise and knowledge work in contexts enabled by extended-context large language models. Drawing on externalist epistemology, ALC reconceptualizes expertise not as the possession of internal knowledge representations but as communicative competence developed through sustained engagement with domain-specific discourse. We propose an hour-logging model of expertise development where measurable time spent in domain communications serves as both a developmental mechanism and a metric for competence. Extended-context LLMs function as mediators that enable continuous access to complete knowledge corpora, fundamentally transforming how individuals and organizations approach learning, expertise development, and knowledge management. We detail organizational and educational implications, propose measurement frameworks, and identify testable propositions. This framework offers an alternative to knowledge graph approaches and traditional knowledge management paradigms, with significant implications for how organizations structure learning, assess expertise, and invest in knowledge infrastructure.
Keywords: expertise development, communicative competence, externalist epistemology, knowledge work, large language models, organizational learning, knowledge management
1. Introduction: From Critique to Construction
The previous papers in this series have systematically dismantled the knowledge graph (KG) paradigm. Paper 1 demonstrated the technical obsolescence of KGs in the face of extended context windows, showing that when you can fit entire knowledge bases into working context, the primary value proposition of structured knowledge graphs—managing bounded contexts—evaporates (Author, 2025a). Paper 2 provided a comprehensive literature review revealing that KG research has failed to address this paradigm shift and harbors unstated epistemological commitments (Author, 2025b). Paper 3 exposed these commitments as rooted in internalist epistemology, which treats knowledge as mental content requiring external structured representation—a philosophical stance increasingly at odds with both cognitive science and the functional capabilities of extended-context AI systems (Author, 2025c).
But critique alone is insufficient. Having identified what is wrong with the dominant paradigm, we must articulate what should replace it. This paper provides that positive vision.
Application Layer Communication (ALC) offers a fundamentally different way of understanding expertise, learning, and knowledge work. Rather than treating knowledge as possessable content requiring extraction, storage, and retrieval, ALC reconceptualizes expertise as communicative competence developed through sustained engagement with domain-specific discourse. This framework aligns with externalist epistemology, where knowledge constitutively depends on environmental and social factors rather than internal representations (Putnam, 1975; Burge, 1979; Clark & Chalmers, 1998). Critically, extended-context LLMs enable this framework to become practically implementable at scale, transforming from philosophical possibility to operational reality.
The stakes are substantial. Organizations invest billions annually in knowledge management infrastructure based on internalist assumptions—knowledge capture programs, expertise location systems, and increasingly, knowledge graphs (Davenport & Prusak, 2000; Nonaka & Takeuchi, 1995). Educational institutions structure curricula around knowledge acquisition and content mastery, measuring success through recall-based assessments (Bloom, 1956; Anderson & Krathwohl, 2001). If these foundational assumptions are misguided, then these investments are not merely suboptimal but potentially counterproductive, focusing effort on the wrong problems entirely.
ALC redirects attention to what actually constitutes expertise: not what someone knows, but how competently they can engage in domain-specific communication. A physicist is expert not because they possess structured representations of physical laws but because they can engage sophisticated discourse about physics—reading papers, formulating problems, critiquing arguments, proposing solutions. An organizational strategist is expert not because they have internalized management frameworks but because they can engage competently in strategic conversations—analyzing situations, evaluating options, articulating recommendations. Extended-context LLMs enable individuals to develop this competence by sustaining rich, hours-long dialogues within complete domain contexts, fundamentally changing what is possible in learning and knowledge work.
This paper proceeds as follows. Section 2 defines Application Layer Communication and its four core components. Section 3 introduces the hour-logging model of expertise, explaining how measurable engagement time serves as both developmental mechanism and competence metric. Section 4 examines how extended-context LLMs enable ALC by mediating access to complete knowledge corpora. Sections 5 and 6 explore organizational and educational implications respectively, showing how ALC reconceives knowledge management and learning. Section 7 proposes measurement frameworks and testable propositions. Section 8 acknowledges limitations and boundary conditions. Section 9 concludes with implications for theory and practice.
2. Defining Application Layer Communication
2.1 Core Concept
Application Layer Communication refers to sustained, domain-specific linguistic engagement that develops facility in accessing distributed knowledge through communicative practice. The term “application layer” derives from networking protocols, where the application layer represents the interface through which users interact with underlying systems. Analogously, ALC represents the communicative interface through which individuals access knowledge distributed across texts, discourse communities, and cultural practices.
Three conceptual pillars support this definition:
First, knowledge is distributed, not internal. Following externalist epistemology (Paper 3; Author, 2025c), knowledge resides in published discourse, ongoing communicative exchanges, institutional practices, and cultural contexts. A medical textbook contains knowledge not because readers internalize its content as mental representations but because competent practitioners can engage it communicatively—cite it, critique it, apply its reasoning, extend its arguments. The knowledge remains distributed across the discourse community; expertise lies in access competence.
Second, expertise is communicative, not possessive. An expert is someone who can engage sophisticated domain-specific communication, not someone who possesses structured knowledge representations. This aligns with Dreyfus and Dreyfus’s (1986) expertise development models, where experts display intuitive, context-sensitive responses rather than rule-following behavior, and with Lave and Wenger’s (1991) situated learning theory, where learning means participating in communities of practice rather than acquiring abstract knowledge.
Third, development occurs through engagement, not acquisition. Expertise develops through hours spent reading domain literature, writing domain-specific content, and engaging in domain discussions. This is not information transmission but linguistic enculturation—learning to speak, read, write, and think within domain-specific communicative norms.
2.2 Four Components of ALC Competence
Application Layer Communication competence comprises four interrelated dimensions:
2.2.1 Domain-Specific Linguistic Competence
This encompasses mastery of:
Terminology: Technical vocabulary, jargon, and their precise usage contexts
Argumentation patterns: How claims are made, supported, and contested within the domain
Genre conventions: How different types of domain communications are structured (research papers, case studies, technical specifications)
Discourse norms: What constitutes valid evidence, acceptable reasoning, and persuasive argument
For example, a competent economist doesn’t just know definitions of “elasticity” or “marginal utility” but understands how these terms function in economic arguments, when their application is appropriate, and how they connect to broader economic reasoning patterns.
2.2.2 Demonstrated Engagement
Measurable time spent in domain-specific communicative activities:
Reading domain literature (papers, books, reports, case studies)
Writing domain content (analyses, proposals, explanations, critiques)
Engaging in domain discussions (meetings, forums, collaborative work)
Interactive work with domain materials (extended context sessions, problem-solving)
This component provides the observable, quantifiable basis for expertise assessment, addressing the persistent challenge of expertise measurement in organizational and educational contexts (Ericsson & Smith, 1991; Shanteau, 1992).
2.2.3 Access Competency
The ability to effectively engage distributed knowledge through communication:
Question formulation: Posing meaningful, well-structured domain queries
Response interpretation: Understanding technical responses, identifying gaps, recognizing misconceptions
Navigation: Moving through complex discourse, following citations, tracing arguments
Synthesis: Integrating information from multiple sources into coherent understanding
This distinguishes superficial engagement from deep competence. A novice and expert may both spend an hour reading the same paper, but the expert’s access competency enables qualitatively richer engagement.
2.2.4 Communicative Agility
Facility in moving between related domains and adapting communicative strategies:
Domain bridging: Connecting concepts across related fields (e.g., applying economic reasoning to organizational behavior)
Audience adaptation: Adjusting technical depth and framing for different interlocutors
Metacommunication: Discussing communication itself, clarifying misunderstandings, negotiating meanings
Rapid contextualization: Quickly establishing shared context in new communicative situations
Communicative agility explains expert adaptability—why domain experts can often engage productively in adjacent fields despite lacking formal training. Their communicative competence transfers across related discourse communities.
2.3 Contrast with Traditional Conceptions
The ALC framework fundamentally reorients how we conceptualize expertise:
Traditional ViewALC ViewExpertise = Knowledge possessionExpertise = Communicative competenceLearning = Acquiring mental contentLearning = Developing linguistic facilityAssessment = Testing recall/understandingAssessment = Evaluating communicative performanceDevelopment = Information transmissionDevelopment = Engagement accumulationExpert systems = Knowledge extraction + storageExpert systems = Communication mediation
This is not merely terminological preference. These conceptual differences drive fundamentally different organizational practices, educational designs, and technology investments. An organization believing expertise equals possession will invest in knowledge capture; one embracing ALC invests in access competence development. A university believing learning equals acquisition designs lectures and tests; one embracing ALC designs sustained engagement and communicative assessment.
3. The Hour-Logging Model of Expertise
3.1 Conceptual Foundation
The hour-logging model posits that expertise develops through and can be meaningfully measured by cumulative time spent in domain-specific communicative engagement. This builds on well-established findings that expertise requires extensive practice (Ericsson, Krampe, & Tesch-Römer, 1993; Gladwell’s popularization via the “10,000-hour rule”), but reconceptualizes what that practice consists of: not acquiring knowledge but developing communicative competence.
Consider software engineering expertise. A developer becomes expert not primarily through reading documentation or attending lectures but through thousands of hours engaging with codebases—reading code, writing code, reviewing code, discussing architectural decisions, debugging problems. These are fundamentally communicative activities: interpreting symbolic structures, expressing intent through code, engaging in technical discourse with team members and documentation. The hours accumulate communicative facility within the domain.
The same pattern holds across knowledge-intensive domains:
Medical expertise: Hours engaging with medical literature, patient cases, clinical discussions, diagnostic reasoning
Legal expertise: Hours reading case law, statutory interpretation, legal argument construction, client consultation
Scientific expertise: Hours with research literature, experimental design, data analysis, scientific writing and review
Strategic expertise: Hours analyzing organizational situations, evaluating strategic options, crafting recommendations
In each case, expertise manifests as facility in domain-specific communication, and development occurs through sustained engagement measurable in hours.
3.2 What Gets Logged
The hour-logging model tracks time spent in specific communicative activities:
Reading Domain Literature
Academic papers, books, technical reports
Industry analyses, case studies, white papers
Historical documents, foundational texts
Quality threshold: Sustained engagement requiring comprehension, not superficial scanning
Writing Domain-Specific Content
Original analyses, research papers, technical specifications
Explanatory documents, teaching materials, documentation
Proposals, recommendations, strategic plans
Quality threshold: Meaningful composition, not mere transcription or templating
Engaging in Domain Discussions
Professional conversations, meetings, consultations
Academic seminars, conference participation
Peer review, collaborative analysis
Online forums and communities (when substantive)
Quality threshold: Active participation, not passive observation
Interactive Work with Domain Materials
Extended-context sessions with LLMs exploring domain knowledge
Problem-solving within domain contexts
Simulation and scenario analysis
Tool-mediated domain exploration
Quality threshold: Genuine inquiry and thinking, not mechanical operations
Importantly, not all time in domain-adjacent activities counts:
Administrative work related to the domain: Does not count
Watching lectures passively: Minimal value (passive consumption vs. active engagement)
Credential-studying for exams: Limited value unless deeply engaging content
Social activities with domain practitioners: Only counts if substantive domain discussion
This selectivity is crucial. The hour-logging model measures communicative engagement specifically, not merely time spent “around” a domain.
3.3 Why Hours Matter as a Metric
Hours logged in communicative engagement offer several advantages over traditional expertise metrics:
Objectivity: Hours are measurable and verifiable, unlike subjective assessments of “understanding” or “mastery.” Organizations can track engagement; individuals can log their development.
Process Alignment: Hours capture the actual developmental process rather than proxy outcomes. Traditional credentials (degrees, certifications) measure participation in credentialing systems, not actual engagement with domain knowledge.
Continuous Measurement: Hours accumulate continuously throughout a career, providing ongoing developmental feedback rather than discrete certification events.
Predictive Validity: While empirical validation is needed (Section 7), the hour-logging model predicts that engagement time should correlate strongly with observable competence—more strongly than traditional metrics in many contexts.
Motivational Clarity: Individuals and organizations gain clear understanding of expertise development: invest hours in substantive engagement. This is actionable in ways that “acquire knowledge” or “demonstrate mastery” are not.
Transfer Transparency: Hours logged across related domains indicate breadth of competence. Someone with 2,000 hours in economics and 1,000 hours in sociology has demonstrable foundation for organizational behavior analysis.
3.4 Thresholds and Development Stages
While the “10,000-hour rule” provides a rough benchmark, ALC recognizes that hour requirements vary by domain complexity and prior foundation:
Foundational Competence (500-1,000 hours): Can engage basic domain discourse, understand common terminology, follow standard arguments. Analogous to Dreyfus and Dreyfus’s (1986) “Advanced Beginner” stage.
Operational Competence (2,000-3,000 hours): Can independently engage domain work, contribute to discussions, identify problems and formulate approaches. Analogous to “Competent” stage.
Expert Competence (5,000-10,000+ hours): Can engage sophisticated domain discourse, evaluate complex arguments, generate novel insights, contribute to domain knowledge. Analogous to “Expert” and “Master” stages.
These thresholds are domain- and context-dependent. A narrow technical specialty may reach operational competence faster than a broad interdisciplinary field. Prior expertise in related domains accelerates development in new domains.
Critically, these stages describe communicative facility, not knowledge accumulation. The 10,000-hour physicist can engage expert-level physics discourse; whether they have “more knowledge” than a 1,000-hour physics graduate student is a category error from the ALC perspective.
3.5 Organizational and Individual Implementation
For organizations, hour-logging enables:
Expertise mapping: Identify who has substantive engagement across domains
Development planning: Structure deliberate expertise cultivation programs
Team composition: Assemble teams based on complementary communicative competencies
Succession planning: Identify gaps in organizational communicative capacity
For individuals, hour-logging provides:
Career development framework: Clear understanding of expertise investment required
Portfolio building: Documented evidence of domain engagement over time
Strategic learning: Conscious allocation of time to high-value communicative engagement
Competence signaling: Alternative to credentials, demonstrating actual development
Implementation requires infrastructure: time-tracking tools integrated with reading, writing, and discussion platforms; organizational norms treating logged hours as legitimate expertise indicators; educational institutions recognizing engagement time in assessment.
4. Extended Context LLMs as ALC Mediators
4.1 Functional Role in Application Layer Communication
Extended-context large language models (LLMs) fundamentally transform ALC from theoretical framework to practical implementation by serving as communication mediators that enable sustained engagement with complete knowledge corpora. Understanding this role requires clarity about what LLMs do and do not do in the ALC framework.
What LLMs Are in ALC: Communication interfaces to distributed knowledge
Enable continuous dialogue within domain contexts spanning millions of tokens
Facilitate natural language engagement with entire knowledge bases
Support multi-hour learning and analysis sessions without retrieval overhead
Mediate access to complete discourse corpora (textbooks, papers, documentation)
Accelerate language learning through interactive, contextualized exposure
What LLMs Are Not in ALC: Knowledge storage or replacement for engagement
Not repositories from which knowledge is extracted
Not substitutes for genuine communicative engagement
Not providers of “quick answers” bypassing development of competence
Not standalone expertise; require ongoing human communicative participation
The distinction is critical. ALC does not treat LLMs as knowledge bases to query but as mediators enabling sustained discourse with domain knowledge. The user develops competence through engagement; the LLM enables that engagement at unprecedented scale and continuity.
4.2 Contrast with Knowledge Graph Approaches
The functional difference between LLMs-as-ALC-mediators and knowledge graphs is fundamental:
Knowledge Graph Paradigm:
Extract knowledge from texts
Structure as entities and relations
Store in graph database
Enable retrieval through queries
Assumption: Knowledge exists as discrete facts requiring external representation
ALC Paradigm with Extended-Context LLMs:
Load complete texts into extended context
Enable natural language engagement with entire corpus
Sustain multi-hour communicative sessions
Develop competence through ongoing discourse
Assumption: Knowledge exists in discourse; expertise is communicative access
Consider a concrete comparison. A student learning machine learning:
KG Approach:
System extracts concepts from ML textbooks and papers
Structures them as graph: “Neural Network” → “Has Layer” → “Dense Layer”
Student queries: “What is backpropagation?”
System retrieves relevant graph nodes and returns answer
Student receives information; communicative engagement minimal
ALC Approach with Extended Context:
Student loads complete ML textbook + seminal papers (within 200K+ token context)
Engages in hours-long dialogue: “Walk me through backpropagation intuition”
LLM draws from complete context to engage explanation, responds to follow-ups
Student explores related concepts, works through examples, poses edge cases
Hours accumulate; communicative competence develops through sustained engagement
The KG approach optimizes for fast answer retrieval; the ALC approach optimizes for sustained communicative engagement that builds competence. These are fundamentally different goals reflecting different epistemological foundations.
4.3 Practical Workflow Examples
Example 1: Policy Analyst Learning Climate Economics
Traditional approach: Enroll in course, read assigned chapters, take exams, receive credential
ALC approach:
Load complete corpus: IPCC reports, climate economics papers (Nordhaus, Stern, etc.), policy analyses
Context: ~150K tokens of core literature
Engage 40-hour learning program:
Hours 1-10: Foundational discourse—understanding climate models, economic frameworks, policy instruments
Hours 11-25: Deep dives—carbon pricing mechanisms, cost-benefit analysis debates, international agreements
Hours 26-40: Synthesis—analyzing specific policy proposals, evaluating trade-offs, developing recommendations
Outcome: Can engage climate economics discourse competently; has 40 logged hours demonstrating engagement
Example 2: Engineer Onboarding to Large Codebase
Traditional approach: Read documentation, pair with senior engineer, gradual task assignment
ALC approach (augmented):
Load complete codebase documentation + architectural decision records + related technical literature
Context: ~80K tokens technical context
Structured 30-hour onboarding:
Hours 1-8: System architecture communication—understanding major components, design patterns, rationale
Hours 9-20: Subsystem deep dives—exploring authentication, data layer, API design
Hours 21-30: Problem-solving scenarios—working through common issues, proposing modifications
Parallel: Pair programming with senior engineers (traditional approach still valuable)
Outcome: Substantially faster ramp-up; logged hours demonstrate systematic engagement
Example 3: Executive Learning Organizational Behavior
Traditional approach: MBA program, case studies, lectures, networking
ALC approach (complementary):
Load corpus: Key OB textbooks, seminal papers (March, Simon, Weick, etc.), Harvard Business Review articles
Context: ~100K tokens OB literature
Ongoing engagement: 2-3 hours weekly over 18 months = ~150 total hours
Topics: Organizational decision-making, culture, change management, leadership, team dynamics
Application: Regular analysis of own organization through OB frameworks during engagement sessions
Outcome: Sophisticated OB communicative competence; documented 150-hour engagement
These examples show how extended-context LLMs enable systematic, scalable ALC implementation. The critical factor is not the technology itself but its use to facilitate sustained, substantive communicative engagement that develops genuine competence.
4.4 Why Extended Context Is Transformative
The jump from 2K to 200K+ token contexts is qualitative, not merely quantitative:
At 2K tokens: Must constantly manage context boundaries; cannot sustain extended exploration; retrieval becomes necessary; KG-like structures provide value
At 200K+ tokens: Can load complete knowledge corpora; enables hours-long coherent dialogues; retrieval optional for many use cases; direct communicative engagement becomes primary mode
This is why extended context changes everything (Paper 1; Author, 2025a). The bounded context problem that necessitated knowledge extraction and structuring disappears. We can engage domain knowledge directly, communicatively, for extended periods—precisely what ALC requires.
Moreover, emerging 2M+ token contexts extend this further, enabling engagement with entire libraries of technical documentation, comprehensive case law databases, or complete research literature for subfields. As context windows expand, the ALC paradigm becomes increasingly practical across more domains and use cases.
5. Organizational Implications: Knowledge Management Reconceived
5.1 From Knowledge Management to Communicative Competence Management
Traditional knowledge management (KM) emerged from internalist assumptions: organizations possess knowledge that must be captured, stored, made accessible, and transferred to employees (Davenport & Prusak, 2000; Nonaka & Takeuchi, 1995). Despite decades of investment, KM initiatives consistently underperform (Storey & Barnett, 2000; Hansen, Nohria, & Tierney, 1999), suggesting fundamental conceptual problems.
ALC reframes the organizational knowledge challenge:
Traditional KM Problem: “How do we get knowledge out of experts’ heads into systems where others can access it?”
ALC Reframing: “How do we develop employees’ communicative competence to access distributed organizational and domain knowledge?”
This reframing shifts organizational focus from knowledge capture to competence development, from building repositories to facilitating engagement.
5.2 Organizational Practices Under ALC
5.2.1 Expertise Tracking and Development
Current Practice: Organizations track credentials (degrees, certifications), job titles, years of experience—all poor proxies for actual competence.
ALC Practice: Track logged hours in domain-specific communicative engagement:
Hours reading technical literature and documentation
Hours writing analyses, specifications, strategic documents
Hours in substantive technical discussions and problem-solving sessions
Hours in extended-context learning and development sessions
Implementation:
Integrated time-tracking with reading, writing, and discussion tools
Regular hour-logging reviews with managers (analogous to current performance reviews)
Individual development plans structured around hour accumulation in strategic domains
Expertise maps showing organizational communicative capacity distribution
Benefits:
Objective competence measurement
Clear development pathways
Identification of expertise gaps
Documentation of actual development vs. credential proxies
5.2.2 Team Composition
Current Practice: Assemble teams based on job functions, credentials, availability
ALC Practice: Compose teams based on complementary communicative competencies:
Required domain competencies for project
Logged hours in each domain by potential team members
Breadth (multiple domains) vs. depth (single domain) trade-offs
Communicative agility across team (ability to bridge domains)
Example: Product development team needs:
User research competence: 2,000+ hours in user research discourse
Technical architecture competence: 3,000+ hours in system design discourse
Market analysis competence: 1,500+ hours in market research discourse
Project management competence: 2,000+ hours in agile/project management discourse
Team assembled by identifying individuals with logged hours in these domains, ensuring communicative capacity exists for all required engagements.
5.2.3 Onboarding and Knowledge Transfer
Current Practice: Shadowing, documentation reading, gradual task assignment, hoping tacit knowledge transfers
ALC Practice: Structured immersion in organizational discourse:
Load new hire with complete organizational documentation, strategic plans, technical materials (via extended context)
Design 80-160 hour onboarding program of systematically engaging organizational knowledge
Track hour accumulation across key organizational domains
Pair extended-context engagement with traditional mentorship (complementary, not replacement)
Assess via communicative performance: Can they engage competently in organizational discourse?
Knowledge Transfer Problem Dissolved: Traditional KM treats knowledge transfer as extracting tacit knowledge from departing experts. ALC recognizes experts have communicative competence, not extractable knowledge. Solution: Accelerate remaining employees’ development of similar competence through intensive engagement with domain discourse (which the expert presumably engaged deeply).
5.2.4 Strategic Learning and Development
Current Practice: Training programs, workshops, credentials—often disconnected from actual work
ALC Practice: Ongoing communicative engagement as core work activity:
Organizations budget employee time for domain engagement (e.g., 10-15% of work time)
Access to extended-context LLM infrastructure for sustained learning sessions
Recognition that expertise development requires actual hours in discourse
Integration of development and work: Learning becomes continuous engagement rather than separate activity
Cultural Shift: From “employees learn then work” to “substantive work includes sustained communicative engagement that develops expertise”
5.3 Investment Reallocation
ALC implies fundamental shifts in organizational technology investment:
Decrease Investment In:
Knowledge graph construction and maintenance
Knowledge extraction projects and tools
Complex knowledge management systems focused on capture/storage
Extensive documentation programs (documentation still needed but emphasis shifts)
Increase Investment In:
Extended-context LLM infrastructure and API access
Time-tracking and engagement logging systems
Protected time for employee communicative engagement
Organizational discourse corpus curation (ensuring quality materials exist)
Communicative competence development programs
ROI Comparison (hypothetical but testable):
KG-based KM: $5M initial investment, $1M/year maintenance, indirect competence benefits
ALC-based KM: $500K extended-context infrastructure, $500K/year operational costs, direct competence development
If ALC approach produces 20% greater employee competence (measurable via performance outcomes), organizational impact is substantial.
5.4 Organizational Learning Reconceived
Organizational learning theory (Argyris & Schön, 1978; March, 1991; Levitt & March, 1988) distinguishes single-loop (adaptive) from double-loop (generative) learning. ALC suggests both occur through communicative processes:
Single-Loop: Employees engage organizational discourse to understand and refine existing practices Double-Loop: Employees engage external discourse (academic literature, industry analyses, alternative frameworks) to question and transform organizational practices
Extended-context LLMs enable both by facilitating sustained engagement with internal and external knowledge sources. Organizations learn as employees develop communicative competencies that enable critical engagement with organizational and domain knowledge.
This connects to absorptive capacity (Cohen & Levinthal, 1990)—organizations’ ability to recognize, assimilate, and apply external knowledge. ALC reframes absorptive capacity as organizational communicative competence: the aggregate capacity of employees to engage external discourse and integrate insights into organizational practice. Developing absorptive capacity means developing employees’ domain-specific communicative competencies.
6. Educational Implications: Learning Reconceived
6.1 From Knowledge Acquisition to Communicative Competence Development
Educational institutions face persistent challenges: students forget content, struggle to apply learning, perform poorly on transfer tasks (Bransford, Brown, & Cocking, 2000). These challenges may stem from fundamentally misconceived learning goals rooted in internalist epistemology.
Traditional Educational Goal: Students acquire knowledge (mental content) ALC Educational Goal: Students develop communicative competence (linguistic facility)
This shift is profound. It changes what we teach, how we teach, how we assess, and what we consider successful learning.
6.2 Pedagogical Practices Under ALC
6.2.1 Sustained Engagement as Core Pedagogy
Traditional: Lectures convey information; readings supplement; assignments test understanding
ALC: Structured immersion in domain discourse with extensive engagement time
Example: Introductory Economics Course
Traditional structure:
30 hours lectures + 60 hours reading + 20 hours problem sets = 110 total hours
Assessment: Exams testing recall and application of concepts
Outcome: Students can answer economics questions on tests
ALC structure:
20 hours introductory lectures establishing domain context
90 hours structured engagement with economics discourse:
Load: Core economics textbook + 10 seminal papers + current economic policy debates
Context: ~60K tokens economic literature
Engagement: Students conduct sustained dialogues exploring economic concepts, working through problems, analyzing policy proposals, critiquing arguments
Facilitation: Instructor monitors engagement, provides challenges, ensures depth
Assessment: Portfolio demonstrating communicative competence (see below)
Outcome: Students can engage economics discourse competently—read economic analyses, critique economic arguments, apply economic reasoning
The ALC version requires similar total time but structures it around active communicative engagement rather than passive information reception.
6.2.2 Authentic Discourse Immersion
Students engage actual domain discourse, not textbook summaries:
Primary sources: Original research papers, foundational texts, professional publications
Professional genres: How practitioners actually communicate in the domain
Current debates: Active controversies and open questions
Multiple perspectives: Competing frameworks and approaches
This aligns with cognitive apprenticeship models (Collins, Brown, & Newman, 1989) and situated learning (Lave & Wenger, 1991), but extended-context LLMs make it scalable. Previously, authentic immersion required small seminars with expert instructors. Now, students can sustain hours-long engagement with complete domain discourse mediated by LLMs, with instructors providing guidance and oversight rather than constant mediation.
6.2.3 Communicative Scaffolding
Extended-context LLMs enable progressive sophistication in discourse engagement:
Beginner Engagement (First 100 hours):
Explicit terminology explanation
Simplified argument structures
Extensive examples and analogies
Patient response to basic misunderstandings
Guided exploration of foundational concepts
Intermediate Engagement (100-500 hours):
More compressed explanations assuming foundation
Introduction of controversies and competing frameworks
Expectation of more sophisticated questions
Less scaffolding, more independent exploration
Advanced Engagement (500+ hours):
Engagement at near-expert discourse level
Critique of arguments, identification of assumptions
Novel synthesis and application
Minimal scaffolding; LLM as peer interlocutor
This progressive sophistication happens naturally through sustained engagement, but can be intentionally designed into educational programs.
6.3 Assessment Reconceived
Traditional assessment measures recall and application through exams, problem sets, papers. These assess internalized knowledge—can students produce correct answers?
ALC assessment measures communicative competence—can students engage domain discourse sophisticatedly?
6.3.1 Engagement Hour Portfolios
What: Documented record of student engagement hours with evidence
Components:
Log of hours spent in reading, writing, discussing domain content
Artifacts demonstrating engagement: Written analyses, discussion transcripts, problem solutions
Reflection on development: How communicative facility has grown
Breadth indicators: Engagement across subdomains or related domains
Assessment Criteria:
Quantity: Total hours (threshold requirements)
Quality: Sophistication of engagement observable in artifacts
Progression: Development arc from beginner to advanced engagement
Authenticity: Genuine inquiry vs. superficial participation
6.3.2 Communicative Performance Tasks
What: Tasks requiring authentic domain discourse engagement
Examples:
Literature Review: Can student engage 20+ papers on topic, identify themes, critique methodologies, synthesize insights? (Not just summarizing—engaging)
Problem Analysis: Can student analyze complex domain-specific problem, apply relevant frameworks, evaluate solution approaches, articulate recommendations?
Debate Participation: Can student argue positions on domain controversies, respond to counterarguments, demonstrate understanding of multiple perspectives?
Peer Review: Can student provide substantive feedback on peers’ work using domain-appropriate criteria?
Assessment: Evaluators (instructors, peers, practitioners) assess communicative sophistication:
Appropriate use of terminology and concepts
Engagement with relevant literature and frameworks
Quality of reasoning and argumentation
Awareness of domain controversies and limitations
Ability to connect across concepts and contexts
6.3.3 Eliminating (Most) Traditional Exams
Multiple-choice and short-answer exams measure recall—precisely what ALC considers misconceived. If knowledge is distributed and expertise is communicative, testing recall of isolated facts assesses the wrong thing.
Exception: Procedural knowledge requiring rapid, correct execution (medical diagnosis, emergency response) where reliability matters more than communicative sophistication. But most higher education contexts do not have this requirement.
Replacement: Communicative performance tasks, engagement portfolios, and crucially, ongoing engagement assessment. Instructors observe students’ developing communicative competence throughout the course via their engagement artifacts, providing feedback that guides development.
6.4 Curricular Implications
6.4.1 Restructuring Around Hour Accumulation
If expertise requires ~10,000 hours and undergraduate education spans ~6,000 class-hours (4 years × 30 hours/week × 50 weeks), students can develop foundational competence in 1-2 domains with operational competence in 2-3 related domains:
Example: Economics Major
Core economics: 3,000 hours → Operational competence
Econometrics/statistics: 1,200 hours → Operational competence
Political economy: 800 hours → Foundational competence
Related domains (history, sociology, business): 600 hours → Foundational competence
General education: 400 hours
This requires ruthless prioritization: depth in few domains rather than superficial coverage of many. The goal is developing actual communicative competence in core domains, not surveying many topics.
6.4.2 Interdisciplinary as Communicative Agility
Interdisciplinary education reconceived: Not combining content from multiple fields but developing ability to engage discourse across related domains. This happens naturally when students have foundational competence in one domain and engage related domains—their communicative competence transfers, enabling faster development in adjacent areas.
Example: Student with 2,000 hours in economics engaging organizational behavior:
Communicative facility with social science reasoning
Understanding of theoretical argumentation patterns
Familiarity with quantitative analysis discourse
Results: Faster development of OB communicative competence than student with no social science background
6.5 Institutional Transformation
Full ALC implementation requires institutional changes:
Infrastructure:
Extended-context LLM access for all students
Tools for logging and demonstrating engagement hours
Faculty development in ALC pedagogical approaches
Assessment systems supporting portfolio and performance evaluation
Cultural:
Accepting engagement hours as legitimate learning metric
Valuing depth over breadth in program design
Reconceiving faculty role from information transmitters to engagement facilitators
Redefining educational success as communicative competence development
Economic:
Less emphasis on lecture halls (economies of scale in information transmission)
More emphasis on engagement facilitation and assessment
Technology investment in LLM infrastructure
Potential cost savings from more efficient competence development
7. Measurement and Validation
7.1 Metrics for ALC Competency
The ALC framework requires measurable indicators of communicative competence across multiple dimensions:
7.1.1 Quantitative Engagement Metrics
Logged Hours:
Primary metric: Total time in domain-specific communicative engagement
Breakdown: Hours in reading, writing, discussion, interactive problem-solving
Distribution: Concentration vs. breadth across subdomains
Trajectory: Hour accumulation rate over time
Engagement Depth Indicators:
Session duration: Length of sustained engagement episodes
Return frequency: Regular vs. sporadic engagement
Topic persistence: Time spent on complex vs. simple topics
Resource range: Number of distinct sources engaged
7.1.2 Qualitative Competence Indicators
Discourse Analysis of Artifacts:
Terminology precision and appropriateness
Argument sophistication and structure
Citation and source engagement patterns
Awareness of domain controversies and limitations
Ability to synthesize across sources
Question Quality Assessment:
Question sophistication (Bloom’s taxonomy levels)
Question independence (genuine inquiry vs. prompted)
Question precision (well-formed vs. vague)
Follow-up quality (building on responses vs. tangential)
Response Interpretation Performance:
Ability to identify gaps or errors in responses
Recognition of when additional information is needed
Capacity to reformulate questions based on responses
Integration of new information with existing understanding
7.1.3 Performance-Based Assessment
Domain-Specific Tasks:
Literature review quality
Problem-solving sophistication
Analytical writing coherence and depth
Presentation clarity and persuasiveness
Peer critique substantiveness
Comparative Performance:
Performance on expert-validated tasks
Peer assessment by domain experts
Ranking within cohort on communicative tasks
Improvement over time on standardized tasks
7.2 Testable Propositions
The ALC framework generates empirically testable predictions:
Proposition 1: Logged engagement hours predict task performance better than traditional credentials (degrees, GPAs) in knowledge-intensive domains.
Test: Longitudinal study comparing hour-logged practitioners vs. credentialed practitioners on domain-specific performance tasks. Control for general cognitive ability and prior experience.
Expected Result: Moderate to strong correlation between hours (>r=0.5) and performance; weaker correlation between traditional credentials and performance after controlling for hours.
Proposition 2: Communicative competence transfers across related domains proportionally to domain similarity.
Test: Measure hour-logged competence in Domain A; assess learning rate in related Domain B vs. unrelated Domain C. Domain similarity defined by discourse overlap (shared terminology, argumentation patterns).
Expected Result: Hours in Domain A accelerate competence development in Domain B but not Domain C. Acceleration proportional to measured discourse similarity.
Proposition 3: Extended-context LLM-mediated engagement produces faster competence development than traditional learning approaches for equivalent time investment.
Test: Randomized controlled trial: Two groups learning same domain content. Group 1: Traditional approach (lectures, readings, assignments). Group 2: ALC approach (extended-context engagement with same material). Equal total time investment. Assessment via communicative performance tasks.
Expected Result: Group 2 shows superior communicative competence on performance tasks despite equal time. Effect size: d ≥ 0.5.
Proposition 4: Organizations with higher aggregate employee hour-logs in strategically relevant domains show superior performance outcomes.
Test: Correlational study across organizations in same industry. Measure aggregate employee hours in domains critical for industry success (e.g., technology companies: software engineering, AI/ML, product management). Assess organizational performance (productivity, innovation metrics, financial performance).
Expected Result: Positive correlation between aggregate hour-logged competence and organizational performance (r ≥ 0.3), controlling for organization size and resource availability.
Proposition 5: Hour accumulation shows diminishing marginal returns with expertise thresholds at approximately 500, 2,000, and 10,000 hours.
Test: Large-scale study measuring performance across hour ranges. Assess marginal performance gains per additional hour at different accumulation levels.
Expected Result: Steeper performance improvement curve in first 500 hours, flatter curve 500-2,000, flatter still 2,000-10,000, minimal improvement beyond 10,000 in many domains.
Proposition 6: Communicative agility (measured by successful engagement in new domains) correlates with prior hour accumulation in related domains.
Test: Measure individuals’ ability to engage novel domain discourse after short immersion (10 hours). Correlate with prior hour accumulation in related domains.
Expected Result: Individuals with 5,000+ hours in related domain show significantly faster development in novel domain than those with <1,000 hours (performance gap at 10 hours: d ≥ 0.7).
7.3 Methodological Considerations
Validating ALC requires addressing several methodological challenges:
Challenge 1: Hour Measurement Accuracy
Solution: Automated tracking where possible; validated self-report protocols; spot-checking and calibration
Challenge 2: Engagement Quality Control
Solution: Artifact analysis to verify substantive engagement; discourse analysis of questions/responses; periodic assessment by domain experts
Challenge 3: Long-Term Longitudinal Data
Solution: Multi-year studies following cohorts; collaboration with educational institutions and organizations committed to hour-logging
Challenge 4: Causal Inference
Solution: Mix of RCTs (where feasible), quasi-experimental designs, and careful controls in correlational studies
Challenge 5: Domain Variation
Solution: Multi-domain studies to test boundary conditions; meta-analysis across domains; theoretical work identifying domain characteristics affecting ALC predictions
7.4 Research Agenda
Comprehensive ALC validation requires sustained research program:
Short-term (1-3 years):
Pilot implementations in educational and organizational settings
Small-scale RCTs comparing ALC vs. traditional approaches
Development of validated hour-logging and assessment tools
Descriptive studies documenting hour accumulation patterns
Medium-term (3-7 years):
Longitudinal studies following cohorts through development
Large-scale organizational implementations with comparison groups
Cross-domain studies testing generalizability
Economic analyses comparing ALC vs. KG/traditional KM ROI
Long-term (7+ years):
Multi-decade studies tracking career outcomes
Generational studies in organizations adopting ALC
Cultural and societal impact studies
Theoretical refinement based on accumulated evidence
This research agenda positions ALC as both practically implementable (organizations and institutions can begin immediately) and theoretically rigorous (subject to empirical validation over time).
8. Limitations and Boundary Conditions
8.1 Where ALC Applies
ALC is most appropriate for:
Knowledge-Intensive Domains with Rich Discourse Communities
Academic disciplines (sciences, humanities, social sciences)
Professional fields (law, medicine, consulting, finance)
Technical specialties (software engineering, data science, architecture)
Strategic and analytical work (policy analysis, organizational strategy)
Contexts Where Expertise = Engagement Ability
Situations requiring analysis, evaluation, synthesis
Work involving interpretation of texts, data, situations
Roles requiring communication of complex ideas
Problems without deterministic solutions requiring judgment
Learning and Development Contexts
Higher education and professional education
Organizational learning and development
Self-directed expertise cultivation
Interdisciplinary and emerging field development
8.2 Where ALC May Not Apply
Procedural Skills Requiring Physical Practice
Surgery, athletics, musical performance, craftsmanship
Skills where communicative understanding is necessary but insufficient
Contexts where bodily knowledge and muscle memory are essential
ALC may support (e.g., understanding anatomy) but cannot replace physical practice
Contexts Requiring Formal Verification
Safety-critical systems (aviation, medical devices, infrastructure)
Regulatory compliance with deterministic requirements
Financial systems requiring audit trails
Situations where “good enough” communicative competence is insufficient; precise, verifiable correctness is required
Highly Standardized Processes
Manufacturing with strict procedural requirements
Rote tasks with minimal judgment requirements
Work following detailed checklists and protocols
Contexts where deviation from standard is error, not expertise
Situations Requiring Real-Time Performance Under Pressure
Emergency response (firefighting, trauma medicine)
Competitive performance (sports, trading)
Contexts where deliberative communicative engagement is too slow
ALC can build foundation, but performance requires additional training
8.3 Open Questions and Future Directions
Question 1: Optimal Hour Distribution Do 10,000 hours distributed as “1 hour daily for 27 years” produce same competence as “40 hours weekly for 5 years”? Spacing effects vs. intensity effects in communicative competence development.
Question 2: Transfer Rates What percentage of hours in Domain A transfer to accelerated development in Domain B? How does this vary by domain similarity, foundational vs. advanced competence, individual differences?
Question 3: Ceiling Effects Do hours matter indefinitely, or does competence plateau? If plateaus exist, where and why? Can targeted interventions extend development curves?
Question 4: Individual Variation Do individuals vary in hour-to-competence conversion rates? If so, what factors (cognitive abilities, learning strategies, prior knowledge) predict variation? Can low converters be helped to improve?
Question 5: Assessment Validity Do hour-based metrics actually predict real-world success better than traditional credentials? Longitudinal career outcome studies needed.
Question 6: Technological Dependence Does ALC create problematic dependence on LLM mediation? Can individuals maintain competence if LLM access is disrupted? What is relationship between mediated and unmediated communicative competence?
Question 7: Quality Control How do we prevent “hour inflation”—logging time without genuine engagement? What mechanisms ensure hour-logging integrity without burdensome oversight?
Question 8: Cultural Adaptation Will hour-logging be culturally acceptable? Do norms around expertise and credentialing resist ALC framing? How can implementation respect existing valuable practices while enabling transformation?
These open questions do not undermine the framework but identify productive research directions and implementation challenges.
8.4 Integration with Existing Practices
ALC should not be viewed as total replacement for all existing practices:
Credentials Still Matter
Signal institutional participation and socialization
Provide structure and sequencing for development
Offer peer communities and networking
ALC complements by adding objective competence measurement
Traditional Assessments Have Roles
Gatekeeping for high-stakes professions (medicine, law)
Ensuring minimum baselines
Regulatory requirements
ALC supplements by providing richer competence picture
Physical and Social Learning Irreplaceable
In-person collaboration and mentorship
Laboratory and fieldwork experiences
Professional socialization and network development
ALC augments by enabling deeper preparation and more productive physical interactions
Organizational Knowledge Practices
Documentation remains valuable (provides discourse to engage)
Mentorship and communities of practice still essential
Experiential learning through doing still critical
ALC enhances by developing competence to engage these more effectively
The goal is not abolishing existing practices but reconceiving their purpose within an externalist, communicative competence framework.
9. Conclusion: A New Paradigm for Knowledge Work
9.1 Synthesis of the Framework
Application Layer Communication offers a fundamentally reimagined understanding of expertise and knowledge work:
Epistemological Foundation: Knowledge is distributed across discourse communities, not possessed as internal content. Expertise is communicative competence enabling access to distributed knowledge, not accumulated mental representations.
Developmental Model: Competence develops through sustained engagement measurable in hours. The hour-logging model provides objective, process-aligned metrics for expertise cultivation and assessment.
Technological Enablement: Extended-context LLMs mediate continuous access to complete knowledge corpora, making sustained communicative engagement practically scalable across domains and contexts.
Organizational Transformation: From knowledge management (capture, storage, retrieval) to communicative competence management (development, assessment, strategic deployment).
Educational Transformation: From knowledge acquisition (content transmission and testing) to competence development (immersive engagement and communicative performance assessment).
This framework coheres conceptually, aligns with externalist epistemology, addresses practical problems in organizational and educational contexts, and generates testable predictions.
9.2 Implications for Theory
For Organizational Theory: ALC reconceptualizes organizational knowledge as aggregate communicative capacity rather than stored content. This connects to resource-based views emphasizing capabilities over assets, but specifies that critical capabilities are communicative. Organizational learning becomes development of collective competence in accessing and engaging relevant discourse. Knowledge transfer is reconceived as accelerated competence development in remaining employees.
For Learning Sciences: ALC provides alternative foundation to cognitivist models emphasizing mental representations. Aligns with sociocultural perspectives (Vygotsky, 1978; Rogoff, 1990) treating learning as participation in communities of practice, but extends through hour-logging operationalization and extended-context enabling technology.
For Philosophy of Mind: ALC instantiates extended mind and active externalism (Clark & Chalmers, 1998) in practical framework. Cognitive processes extend beyond brain/body into linguistic artifacts and communicative practices. Extended-context LLMs become cognitive prosthetics enabling access to distributed cognitive resources.
For Educational Philosophy: Challenges banking model of education (Freire, 1970) where knowledge is deposited into students. Proposes communicative engagement model where students develop linguistic facility in domains, aligning with constructivist and social-constructivist pedagogies but operationalizing them through hour accumulation and extended-context mediation.
9.3 Implications for Practice
For Organizations:
Audit current knowledge management investments: Are they solving the right problem?
Pilot hour-logging programs for critical expertise domains
Invest in extended-context LLM infrastructure enabling employee engagement
Restructure development programs around sustained communicative engagement
Begin measuring expertise through engagement hours alongside traditional credentials
For Educational Institutions:
Experiment with ALC pedagogies in courses where applicable
Develop hour-logging and communicative assessment infrastructure
Restructure curricula prioritizing depth over breadth
Invest in extended-context LLM access for students and faculty
Conduct validity studies comparing ALC vs. traditional outcome metrics
For Individuals:
Begin logging hours in domains of strategic career interest
Use extended-context LLMs for systematic domain immersion
Build portfolios demonstrating communicative competence development
Recognize that expertise development requires actual hour investment
Prioritize depth of engagement over breadth of superficial exposure
9.4 A Vision of Transformed Knowledge Work
Imagine an organization in 2030 where:
Employees routinely engage 3-5 hour extended-context sessions deeply exploring domain knowledge relevant to current challenges
Expertise is mapped through hour-logs showing who has substantive communicative capacity in which domains
Teams are composed by identifying required communicative competencies and matching to individuals’ logged capabilities
Development plans target specific hour accumulations in strategic domains with allocated organizational time
New hires immerse in organizational discourse through structured 100+ hour engagement programs
Success is measured partly by aggregate organizational communicative capacity growth
Imagine an educational institution in 2030 where:
Students engage 20-30 hour weekly immersions in domain discourse via extended-context systems
Courses are structured around systematic engagement with primary literature and professional discourse
Assessment emphasizes communicative performance demonstrated through portfolios and authentic tasks
Degrees certify documented hours in domains alongside communicative competence demonstrations
Graduates emerge with logged 3,000-5,000 hours in core domains, prepared for sophisticated professional engagement
These visions are not utopian fantasies but practical possibilities enabled by extended-context LLMs and reconceptualization of knowledge work through ALC. Implementation requires philosophical reconception, technological investment, cultural adaptation, and institutional transformation—significant but achievable changes.
9.5 Final Reflection
Papers 1-3 of this series demolished the knowledge graph paradigm: showing its technical obsolescence (Paper 1), exposing gaps in supporting literature (Paper 2), and revealing flawed epistemological foundations (Paper 3). This paper completes the series by offering the alternative: a framework that aligns with externalist epistemology, leverages extended-context capabilities, and provides practical guidance for organizations and educational institutions.
Application Layer Communication is not merely a technological shift but a paradigmatic reconceptualization of knowledge work. It requires questioning deeply held assumptions about knowledge, expertise, learning, and organizational capability. Such fundamental questioning is uncomfortable but necessary.
The knowledge graph era is ending not just because extended contexts make them technically unnecessary, but because they were solving the wrong problem. The problem was never “how do we structure and store knowledge?” The problem is “how do we develop communicative competence enabling access to distributed knowledge?” ALC answers that question.
The implications are profound. Billions invested in knowledge management infrastructure may need redirection. Educational practices structuring billions of student-hours may need reconception. Professional credentialing systems may need supplementation or replacement. But the potential gains—more effective learning, more capable organizations, more meaningful measurement of expertise—justify the disruption.
The future of knowledge work is not about building better knowledge graphs. It is about developing communicative competence through sustained engagement with domain discourse, mediated by extended-context systems that make such engagement practically scalable. That future is available now for organizations, institutions, and individuals willing to embrace it.
Application Layer Communication offers the framework. The implementation remains ahead.
References
Anderson, L. W., & Krathwohl, D. R. (Eds.). (2001). A taxonomy for learning, teaching, and assessing: A revision of Bloom’s taxonomy of educational objectives. Longman.
Argyris, C., & Schön, D. A. (1978). Organizational learning: A theory of action perspective. Addison-Wesley.
Author. (2025a). The death of knowledge graphs: Why extended context windows change everything. arXiv preprint.
Author. (2025b). Knowledge graphs: A critical literature review through the lens of extended context. AI Magazine.
Author. (2025c). The epistemological foundations of knowledge representation: Why knowledge graphs get knowledge wrong. Minds and Machines.
Bloom, B. S. (1956). Taxonomy of educational objectives: The classification of educational goals. Longman.
Bransford, J. D., Brown, A. L., & Cocking, R. R. (Eds.). (2000). How people learn: Brain, mind, experience, and school. National Academy Press.
Burge, T. (1979). Individualism and the mental. Midwest Studies in Philosophy, 4(1), 73-121.
Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58(1), 7-19.
Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-152.
Collins, A., Brown, J. S., & Newman, S. E. (1989). Cognitive apprenticeship: Teaching the crafts of reading, writing, and mathematics. In L. B. Resnick (Ed.), Knowing, learning, and instruction: Essays in honor of Robert Glaser (pp. 453-494). Lawrence Erlbaum Associates.
Davenport, T. H., & Prusak, L. (2000). Working knowledge: How organizations manage what they know. Harvard Business School Press.
Dreyfus, H. L., & Dreyfus, S. E. (1986). Mind over machine: The power of human intuition and expertise in the era of the computer. Free Press.
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363-406.
Ericsson, K. A., & Smith, J. (Eds.). (1991). Toward a general theory of expertise: Prospects and limits. Cambridge University Press.
Freire, P. (1970). Pedagogy of the oppressed. Continuum.
Hansen, M. T., Nohria, N., & Tierney, T. (1999). What’s your strategy for managing knowledge? Harvard Business Review, 77(2), 106-116.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge University Press.
Levitt, B., & March, J. G. (1988). Organizational learning. Annual Review of Sociology, 14, 319-340.
March, J. G. (1991). Exploration and exploitation in organizational learning. Organization Science, 2(1), 71-87.
Nonaka, I., & Takeuchi, H. (1995). The knowledge-creating company: How Japanese companies create the dynamics of innovation. Oxford University Press.
Putnam, H. (1975). The meaning of “meaning”. Minnesota Studies in the Philosophy of Science, 7, 131-193.
Rogoff, B. (1990). Apprenticeship in thinking: Cognitive development in social context. Oxford University Press.
Shanteau, J. (1992). Competence in experts: The role of task characteristics. Organizational Behavior and Human Decision Processes, 53(2), 252-266.
Storey, J., & Barnett, E. (2000). Knowledge management initiatives: Learning from failure. Journal of Knowledge Management, 4(2), 145-156.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press.