The motivation for this discussion is to look at the cognitive and logical rationales of e-Learning objects which reside in computer-based e-Learning artefacts. e-Learning objects, the system to which they belong, and the sequence of messages that form a discourse between the system and its environment are the topic of discourse and are inseparable. Altogether, they formulate the Universe of Discourse (, p. 14). When we talk about systems, we equally refer the discourse to e-Learning objects because they are the “workers” of the system. e-Learning is “a combination of content and instructional methods delivered by media elements such as words and graphics on a computer intended to build job-transferable knowledge and skills linked to individual learning goals or organizational performance” (, p. 311). The sciences of instruction, learning, and knowledge are intricate and the “e-” before “Learning” adds another dimension of complexity while paving new learning paths for e-Learning.
e-Learning involves multiple disciplines e.g., philosophy, psychology, pedagogy, anthropology, artificial intelligence (e.g., Artificial Intelligence in Education (AIED)), and human computer interaction (HCI) . e-Learning artefacts should be more than just a technical solution such as for instance a web-based e-Learning site (however sophisticated it may be) containing stylish multimedia assets, Java applets, and dynamic database bindings.
The user range interacting with e-Learning artefacts is large and multifaceted, too. Main actors are probably pedagogues, instructional designers, psychologists, and learners. Not only do they have their individual expectations and assumptions towards an e-Learning artefact but also hold varying degrees of proficiency and professionalism (e.g., not every educator is a professional pedagogue; not every psychologist is a professional pedagogue), motivation of learning (e.g., an adult student may be more motivated to learn than a student who has to memorise chemical formulas for an examination; an educator may not be motivated to be taught by a computer system), education, environment, and alike. Hence, a non-expert may expect or assume a system’s expert intelligence while an expert may wish to choose from available options or templates within the system, to adapt to, or even evolve further from behaviour (i.e., referring to knowledge representation, user behaviour, user action; cf. ). Similarly, a student may expect a virtual instructor to interact alike an educator in the physical world. Accordingly, the attribution of action by users effects an immediate machine reaction . Although we are not exploring the technical aspects of interaction itself (cf. , ), we are interested in the contextual causes by entities, actors, events, and requirements. Hence, expectations, assumptions, intended effects, underlying plans, situated actions , observable, and unobservable behaviour of ourselves and those of others must be known, distinguished, and formulated into explicit requirements to design e-Learning artefacts. However, observing processes is unnatural for us. These processes contain situated actions and reactions as shown in the following example: “… one shopper found an unusually high priced package of cheese in a bin. He suspected an error. To solve the problem, he searched through the bin for a package weighing the same amount and inferred from the discrepancy between the prices that one was in error. His comparison with other packages established which was the errant package. Had he not transferred the calculation to the environment, he would have had to divide weight into price, mentally, and compare the result with the price per pound printed on the label, a much more effortful and less reliable procedure” (, p. 77). In , the authors conclude on their example that the store setting and activities within the store "mutually create and change each other”. Stimuli for such changes in our physical world are situations, or as the authors in  argue: “Situations might be said to co-produce knowledge through activity. Learning and cognition, it is now possible to argue, are fundamentally situated”.
e-Learning artefacts are not only to deliver but also to build job-transferable knowledge and skills so that e-Learning systems, in particular more than any others, should ultimately accomplish man-like behaviour, building on “the notion of a self-explanatory artefact … In this … sense the goal is that the artefact should not only be intelligible to the user as a tool, but that it should be intelligent – that is, able to understand the actions of the user, and to provide for the rationality of its own” (, p. 17). Consequently, e-Learning artefacts should be intelligent with clear-set learning goals i.e., "... the focus in thinking about distributed intelligence is not on intelligence as an abstract property or quantity residing in either minds, organizations or objects. In its primary sense here, intelligence is manifested in activity that connects means and ends through achievements" .
While studies of mutual intelligibility have been concerned with human action, we now have a technology available in e-Learning that has brought with it the idea rather than just using machines, we interact with them as well . Such interaction needs to investigate and separate the study of mutual intelligibility: The relation between observable behaviour and the processes not available to direct observation, that make behaviour meaningful . The expression “mutual” refers to and includes numerous actors and their disparate levels of interaction with e-Learning artefacts while processes of behaviour are essentially cognitive. The gain of interaction is the value of the perspective of one other and is a primary learning constituent in constructivist learning theories , inducing mindfulness in learners .
In view of most computer systems, however, we assume that a plan determines purposeful action  whereof “A plan is any hierarchical process in the organism that can control the order in which a sequence of operations to be performed” (, p. 17). This rational notation also connotes the accord of plans and goals, past actions, effects, their pre- and post-conditions, alternative and future actions.
The domain of systems has thus revealed the importance of the psychological (i.e., cognitive) and social rationales. These must be reflected in e-Learning systems more than in any other computer-based systems because learning is of psychological and pedagogical nature and contains psychological and pedagogical norms. Entities of e-Learning artefacts are primarily pedagogues and learners and their communication is subject to laws of psychology and pedagogy. Taken from here, we must precisely identify tasks and responsibilities of an e-Learning system and decompose them to analyse the functions, communication, and behaviour of e-Learning objects. Their authentic “... tasks and content analysis should focus less on identifying and prescribing a single, best sequence for learning and more on selecting tasks that are both meaningful and able to accomodate constructivistic applications” (, pp. 29-30). The answer to such crucial information can be dynamically inferred from its environment . The key to such information is the subject domain (as discussed in the subsequent section). Notably, such context-driven, evaluative information is meaningful although "… real-world criteria may be very objective. But they are real-world, and to the extent that they reflect real-world criteria, they are meaningful" (, p. 30).
The subject domain (also known as the Universe of Discourse, ) helps us to capture contextual actors, messages, responsibilities, and alike: “So the subject domain of the system is the part of the world that the messages received and sent by the system are about. To find out what the subject domain of a system is, ask what entities and events the messages sent and received by the system are about … to count as elements of the subject domain, these entities and events must be identifiable by the system” (, p. 16). The author in  further specifies, that “The subject domain of a system not only consists of nature and previously installed systems … but also of people and their socially defined reality, including norms and meaning conventions” (, p. 2). The collection of all possible symbolic interactions is called functionality and consists of three classes i.e., the information function, the control function, and the declarative function . Functionality of a system is achieved by its objects and components. Therefore, when we refer to a system in the following discussion, we must bear in mind that it is the responsibility of its objects and components to realise the functionality.
We can now induce that the subject domain of intelligent e-Learning artefacts talks about delivering and constructing pedagogically and psychologically valid learning contents to learners. Therefore, messages (e.g., pedagogical, psychological, and alike) must be learner-centred (, , ) and not instructor-centred while learner-centred-ness also includes autonomy and control . On the contrary, if we assumed that an intelligent e-Learning system should help and instruct educators design pedagogically and psychologically valid e-Learning contents, the subject domain then talks about the construction of contents. In this case, messages are still learner-centred because the educator becomes a type of learner. Also, if both foci fell into place learning activities would truly be blended (cf. ). In either case, however, the subject domain of a system always talks about identifiable methods and events to construct, build, and deliver learning contents to learners but not about the actual learning contents itself. The exclusion of contents supports Moore’s argument of the three types of interaction (i.e., Learner-Content Interaction, Learner-Instructor Interaction, Learner-Learner Interaction ) because the subject domain talks about how learners acquire intellectual facts and not about the contents of the intellectual facts.
The nature of a learner-centred interaction is something much greater than a simple transmission of information, navigating through learning contents , or “a mere process of passive reception and acquisition of knowledge” (, p. 497; ). We would therefore expect these types of events and messages of the subject domain to be of pedagogical nature to facilitate learning. Nevertheless, pedagogy is rather heuristic, an objective experience of how to teach, and is henceforth primarily derived from situated actions. Pedagogy is argued to be ill-structured , neglects research of theories in educational technology , and is thus not suitable for computer-based artefacts which are built on planned actions and intents contrary to situated experience. Situated actions in pedagogy become problematic when it comes to designing e-Learning artefacts so that this, in turn, exactly becomes one of our greatest challenges. According to , one of the propositions of the ethnomethodological view of purposeful action and shared understanding claims that plans are representations of situated actions. Rather than direct situated action, rationality anticipates action before the fact, and reconstructs it afterwards (, p. 53; ). However, as of today, no such e-Learning architectures exist, yet. Contrary to the pedagogical predicament, theories originating from psychology (e.g., instructional processing theory, instructional-design theory ) tell us how knowledge is represented, built, processed and alike in memory. In , the authors argue that instruction is "A deliberately arranged set of external events designed to support the learning process" (, p. 11). Instructional events are "… external, when deliberately planned and arranged constitute instruction" (, p. iv). Although there exists a myriad of psychological learning theories, models, and principles from cognitive and constructivist psychology, precise methodologies are needed specifically for e-learning which allow objects of a system to execute identifiable events and messages. Moreover, such events and messages must be based on non-contradicting approaches as for example, an instructivist versus a constructivist approach. Beyond, with a complex system in mind where for instance the artefact instructs an educator to construct a course and also delivers learning contents to a learner, learning processes cannot be fixed. Hence, our challenge as designers is to find both an optimal learning and teaching process, and identify how to support the interaction between them. To assume that a singular learning process will suffice may be a grave error in design.
Entities of the subject domain are reliant on the type of e-Learning system. We will find a broad modality  spectrum such as mobile learning, web-based learning, distance learning, and more. Geographical distance is less important than the interaction between the learner and the educator (, , ). As discussed earlier, the system could be the educator itself which supports our argument that e-Learning artefacts must be intelligent as well as pedagogically and psychologically valid, more than most other systems. Presently, the most popular e-Learning systems deliver online packaged or instructor-led (i.e., system-led or man-led) courses and tutorials so that we infer that packaged courses or tutorials led by the system have more responsibilities than instructor-led ones. In either case, the subject domain talks about the nature and norms of an entire course structure which is the composition of its individual components (e.g., module, lesson, assignment). The actual learning contents which is outside the responsibility of the subject domain is the substance of the individual components. The term component used here is not to be confused with Merrill’s Component Display Theory (CDT) (, ) although it is interesting to note how the CISCO course structure  applies Merrill’s components. Even though the subject domain talks about how to construct a course, no rational, pedagogical models exist that would tell us how course structures, lectures, or components are to be built validly. The term “lecture” hereto implies cognitive and constructivist learning processes and differentiates from a conventional, instructivist lecture as known from the physical world. At a lower level, we will find principles of instructional design  and cognitive learning principles (e.g., modality principle, contiguity principle; cf. , , ). Despite their cognitive values instructing us of how to reduce the burden of working memory, these principles however, are nearly impossible for a system to put into practice. For example, these principles teach us that presenting visuals emphasising relevant and critical details is effective while arbitrarily adding visuals does not increase learning at all. Henceforth, a system will not be able to elaborate when a picture will be too many. Again, it is the educator or the instructional designer who holds responsible for the contents. But, what about if the system is to help or instruct the educator or instructional designer? The issue is relevant because often, educators believe that fine-looking pictures make a lesson look attractive or give a relaxing ambience and do not know that learning is better when extraneous, content irrelevant materials are excluded (contiguity principle).
At last, the subject domain must therefore talk about logical operability within course structures, the binding between contents based on logical acquaintances and aggregations. These include knowledge mining, functional operability and computational algorithms, and many more to enable the system’s functionality.
Commonly known definitions of e-Learning objects clearly lack the significance of our previous discussion based on the norms and contextual nature of the systems in which e-Learning objects reside. Known definitions do not take intelligence and pedagogical and psychological validity into account, miss the responsibilities they have to fulfil, and the complex interactions they must accomplish. For example, learning or e-Learning objects known as of today take observation from an instructivist view point (e.g., the Lego metaphor and its debate; cf. , , ), mix the physical learning with the digital “e-” learning world , are unclear about granularity , lifespan , and nature (e.g., database entities ).
However, the stringent requirements imposed by the subject domain can best be accomplished by an object-oriented (OO) approach. An OO paradigm offers better support for reuse, is qualified to incorporate instructional operations, and outweighs the capabilities of static database entities. Objects as known by the OO paradigm can be assigned responsibilities, change internal states, communicate by exchanging messages, react to external stimuli, cause effects, and respond to internal causes.
e-Learning artefacts, their objects, and the discourse they have with their environment express an e-Learning based Universe (of Discourse). While the domain of learning itself is already complex by itself, the technological edge in e-Learning increases the intricacy. Yet, technology presents new dimensions in learning. e-Learning includes various disciplines which call for the need of an intensified communication and research when it comes to devising e-Learning artefacts. Their range of users is versatile, holding different expectations and assumptions towards systems, portraying various degrees of proficiency and professionalism, learning motivation, levels of education, and more. Human processes and interactions are intertwined and rely on underlying plans, situated actions, observable and unobservable behaviour, and alike. Humans learn and co-produce knowledge by these processes and interactions. Similarly, e-Learning artefacts which are to build job-transferable knowledge and skills linked to learning goals must be intelligent because they should be able to understand human processes and interactions.
Traditionally, the learning domain has been a highly complicated, meaningful process and interaction between an educator and a learner. With e-Learning artefacts, we now not only find another intelligent actor in the picture, add new degrees of interaction but moreover need to introduce new levels of processes and behaviour that are essentially cognitive, inducing mindfulness in learners. e-Learning artefacts are primarily built on planned action and are subject to the nature, norms, and laws of pedagogy and psychology. From this environment we can dynamically infer tasks and responsibilities for e-Learning artefacts. To comprehend the notion of the subject domain and identify the part of the world that is relevant to a situational e-Learning artefact, we have chosen the metaphor that the system simply exchanges messages with its environment. This functional decomposition must result in identifiable entities, messages, and events to the system.
Pedagogical and psychological construction and delivery of contents rather than the actual content are major key issues. Thereupon, intelligent e-Learning artefacts must be entirely learner- rather than instructor-centred to the one, and prove pedagogical and psychological validity to the other.
Learner-centred interaction in e-Learning is about actively constructing knowledge in a learner’s memory. This is a shift of the paradigm from the conventional way of instruction, where the learner is reckoned a passive vessel ready to be filled with knowledge. Messages and events of the subject domain are therefore both concerned with the delivery of content and active knowledge construction. The interactive and constructive process to deliver learning content embraces a multitude of facets and possible scenarios e.g., the roles, types, and levels of interactions between the system, the instructor, and the learner. Intuitively, we would assume that such delivery processes should primarily be of pedagogical nature but because pedagogy is mainly based on situated experience it is not suitable for e-Learning systems. On the contrary, psychological learning theories, models, and principles seem to give way to deliver learning content but they are still not sufficiently refined yet to be identified by a system. Also, these theories, models, and principles must all cohere within an artefact. Likewise, learning processes cannot be fixed and we are challenged to find optimal learning and teaching processes and to know the interaction between them. To assume that a singular learning process would suffice may be a huge error in design. Another particular area of messages and events of the subject domain relates to e.g., logical operability within course structures, knowledge mining, and computational algorithms.
De facto, presently known definitions of learning objects for e-Learning artefacts have not taken the diverse cognitive and logical rationales into account which primarily come from the domain of learning based on pedagogy and psychology. More detailed, the subject domain has helped us decompose entities, events, messages, laws, nature, norms, and alike so that we have posited that an object-oriented approach would be most suitable to satisfy these stringent requirements. Although the “e” before learning has brought forth new complexities of interaction, mutual intelligibility, situated action, planned action, pedagogical and psychological validity the “e” is maybe also about to bring forth a new shift of paradigm to conventional learning. We are challenged to re-evaluate at how humans perceive computer-based e-Learning artefacts, design learner-centred e-Learning systems and their objects while benefiting from the multiple disciplines involved in e-Learning, resolve semantic spaces between pedagogical and psychological learning theories, models, and principles to make them identifiable for computer-based systems, and to construct teaching models where the educator, the system, and the learner evolve together.
 R. J. Wieringa, Design Methods for Reactive Systems: Yourdon, Statemate, and the UML, Morgan Kaufmann, San Francisco, CA, USA, 2003.
 R. C. Clark and R. E. Mayer, e-Learning and the Science of Instruction, John Wiley, Hoboken, 2002.
 K. Issroff and E. Scanalon, "Educational Technology: The Influence of Theory", Journal of Interactive Media in Education, no vol. (2002) 6, pp. 1-11.
 A. J. Dix, J. E. Finlay, G. D. Abowd, and R. Beale, Human-Computer Interaction, Prentice Hall, Harlow, 1998.
 S. Turkle, The Second Self: Computers and the Human Spirit, Simon and Schuster, New York, 1984.
 B. Muirhead and C. Juwah, "Interactivity in computer-mediated college and university education: A recent review of the literature", http://ifets.ieee.org/discussions/discuss_november2003.html, 2003.
 A. Repenning, A. Ioannidou, and J. Ambach, "Learn to Communicate and Communicate to Learn", Journal of Interactive Media in Education, 98 (1998) 7, pp. 1-50.
 L. A. Suchman, Plans and Situated Actions: The Problem of Human Machine Communication, Cambridge University Press, Cambridge, 1987.
 J. Lave, M. Murtaugh, and O. de la Rocha, "The Dialectic of Arithmetic in Grocery Shopping", in Everyday Cognition: Its Development in Social Context, B. Rogoff and J. Lave, Eds. Cambridge (MA): Harvard University Press, 1984, pp. 67-94.
 J. S. Brown, A. Collins, and P. Duguid, "Situated Learning and the Culture of Learning", Education Researcher, 18 (1989) 1, pp. 32-42.
 R. D. Pea, "Practices of distributed intelligence and design for education", in Distributed cognitions: Psychological and Educational Considerations, G. Salomon, Ed. Cambridge (MA): Cambridge University Press, 1993, pp. 47-87.
 D. H. Jonassen, "Evaluating constructivistic learning", Educational Technology, XXXI (1991) 9, pp. 28-33.
 E. J. Langer, Mindfulness, Perseus Publishing, Cambridge (MA), 1990.
 G. A. Miller, E. Galanter, and K. H. Pribram, Plans and the Structure of Behavior, Holt, Rinehart and Winston, New York, 1960.
 R. J. Wieringa, "The Declarative Problem Frame: Designing Systems that Create and Use Norms", presented at The Tenth International Workshop on Software Specification and Design (IWSSD'00), San Diego, 2000.
 R. Moreno and R. E. Mayer, "A Learner-Centered Approach to Multimedia Explanations: Deriving Instructional Design Principles from Cognitive Theory", Interactive Multimedia Electronic Journal of Computer-Enhanced Learning, 2 (2000) 2, pp. http://imej.wfu.edu/articles/2000/2/05/index.asp.
 D. Smulders, "Designing for Learners, Designing for Users", http://elearnmag.org/subpage/sub_page.cfm?section=3&list_item=11&page=1, 2003.
 P. G. Coman, "Critical success factors for eLearning delivery", presented at Computers in Education, 2002. Proceedings. International Conference on Computers in Education, 2002.
 F. Saba and R. L. Shearer, "Verifying Key Theoretical Concepts in a Dynamic Model of Distance Education", The American Journal of Distance Education, 8 (1994) 1, pp. 36-59.
 M. Orey, "Definition of Blended Learning", http://www.arches.uga.edu/~mikeorey/blendedLearning/, 2003.
 M. G. Moore, "Three Types of Interaction", The American Journal of Distance Education, 3 (1989) 2, pp. 1-6.
 M. B. Nunes and M. McPherson, "Constructivism vs. Objectivism: Where is difference for Designers of e-Learning Environments?" presented at The 3rd IEEE International Conference on Advanced Learning Technologies (ICALT'03), Athens, 2003.
 G. Giebler, "Objectivism to Constructivism: Transforming the Model", Journal of Educational Computing, Design & Telecommunications, 1 (2000) 1, pp. http://coe.ksu.edu/jecdt/Vol_1/articles/giebler.html.
 H. Allert, H. Dhraief, and W. Nejdl, "Meta-Level Category 'Role' in Metadata Standards for Learning: Instructional Roles and Instructional Qualities of Learning Objects", presented at The Second International Conference on Computational Semiotics for Games and New Media (COSIGN 2002), Augsburg, 2002.
 G. H. Mead, "Mind, Self and Society from the Standpoint of a Social Behaviorist". Chicago: University of Chicago Press, 1934.
 C. M. Reigeluth, Instructional-Design Theories And Models: A New Paradigm of Instructional Theory, Lawrence Erlbaum, London, 1999.
 R. M. Gagné, L. J. Briggs, and W. W. Wager, Principles of Instructional Design, Wadsworth/Thomson Learning, Belmont, 1992.
 R. M. Gagné, The Conditions of Learning and Theory of Instruction, Wadsworth, New York, 1985.
 brandonhall.com, "New Technology Definitions", http://www.brandonhall.com/public/glossary/glossary.html, 2003.
 M. G. Moore, "Toward a Theory of Independent Learning and Teaching", Journal of Higher Education, 44 (1973) 9, pp. 661-679.
 N. W. Coppola and R. Myre, "Corporate Software Training: Is Web-Based Training as Effective as Instructor-Led Training?" IEEE Transactions on Professional Communication, 45 (2002) 3, pp. 170-186.
 M. D. Merrill, Component Display Theory, Lawrence Erlbaum Associates, Mahwah, 1983.
 R. C. Clark, Developing Technical Training, Book News, Washington, 1999.
 CISCO, "Reusable Learning Object Authoring Guidelines: How to Build Modules, Lessons, and Topics", http://business.cisco.com/servletwl3/FileDownloader/iqprd/104119/104119_kbns.pdf, 2003.
 R. Clark and G. L. Harrelson, "Designing Instruction That Supports Cognitive Learning Processes", Journal of Athletic Training, 37 (2002) 4 suppl., pp. S-152 – S-159.
 A. Littlejohn, "Reusing Online Resources: A Sustainable Approach to eLearning", Journal of Interactive Media in Education, 2003 (2003) 1, pp. 1-7.
 D. Rehak and R. Mason, "Chapter 3: Keeping the Learning in Learning Objects", Journal of Interactive Media in Education, 2003 (2003) 1, pp. 1-5.
 LTSC, "Draft Standard for Learning Object Metadata", http://grouper.ieee.org/p1484/wg12/files/LOM_1484_12_1_v1_Final_Draft.pdf, 2002.
 D. A. Wiley, The Instructional Use of Learning Objects, Agency for Instructional Technology, Bloomington, 2002.
 J. J. L'Allier, "NETg's Map to Its Products, Their Structures and Core Beliefs", http://www.netg.com/research/whitepapers/frameref.asp, 1997.
 M. D. Merrill, "Knowledge Objects", http://www.id2.usu.edu/Papers/KnowledgeObjects.PDF, 1998.
 IMS Global Learning Consortium, "IMS Learning Design Specification", http://www.imsglobal.org/learningdesign/index.cfm, 2004.
 G. G. Passmore, "Using Vee Diagrams To Facilitate Meaningful Learning and Misconception Remediation in Radiologic Technologies Laboratory Education", Radiologic Science & Education, 4 (1998) 1, pp. http://www.aers.org/V4N1PASSMORE.HTML.
 J. D. Novak and D. B. Gowin, Learning How to Learn, Cambridge University Press, Cambridge, 1984.
 J. D. Novak, "The Theory Underlying Concept Maps and How To Construct Them", http://cmap.coginst.uwf.edu/info/, N/A.
 A. J. Cañas, K. M. Ford, J. D. Novak, P. Hayes, T. R. Reichherzer, and N. Suri, "Using Concept Maps with Technology to Enhance Collaborative Learning in Latin America", http://www.ihmc.us/users/acanas/Publications/QuorumSoupST/SoupsST.htm, N/A.
 Institute for Human and Machine Recognition, "Research :: The Bayesian Network Lens", http://www.ihmc.us/research/projects/BayesLens/, N/A.
 J. Mintzes, J. Wandersee, and J. Novak, Teaching Science for Understanding : A Human Constructivist View, Academic Press, San Diego, 1998.
 F. Van Assche, "Global e-Learning Summit: Standards at Work: Examples from European Schoolnet", http://www.adlnet.org/index.cfm?fuseaction=rcdetails&libid=656&bc=false, 2004.
 J. Tholander and K. Karlgren, "Support for Cognitive Apprenticeship in Object-Oriented Model Construction", http://newmedia.colorado.edu/cscl/220.pdf, 2002.