Bridging Learning Theories
and Technology-Enhanced
Environments: A Critical Appraisal
of Its History
Joost Lowyck

In education, retrospection is often used as a method for better understanding emerging
trends as documented in many books and articles. In this chapter, the focus is not on a broad
description of the history of educational technology but on the interplay between learning
theories and technologies. However, neither learning theories nor tools are monolithic
phenomena. They are composed of multiple attributes, and they refer to many aspects and facets which render the history of educational technology highly complex. Moreover,
evolution in both theory and technology re fl ects no clear successive breaks or discrete
developments—rather, waves of growth and accumulation. When looking closer at learning
and technology, it becomes clear that many interactions occur. These interactions will be
documented following continuous development after World War II. We do not follow a
strict timeline but cluster the critical appraisal in the following observations: (1) evolutions
in society and education have in fl uenced the selection and use of learning theories and tech-
nologies; (2) learning theories and technologies are situated in a somewhat vague concep-
tual fi eld; (3) learning theories and technologies are connected and intertwined by
information processing and knowledge acquisition; (4) educational technologies shifted
learner support from program or instructor control toward more shared and learner control;
and (5) learning theories and fi ndings represent a fuzzy mixture of principles and applications.
The history re fl ects an evolution from individual toward community learning, from content-
driven learning toward process-driven approaches, from isolated media toward integrated
use, from presentation media toward interactive media, from learning settings dependent on
place and time toward ubiquitous learning, and from fi xed tools toward handheld devices.
These developments increasingly confront learners with complexity and challenge their
responsibility to become active participants in a learning society. Keywords
Learning theories • Educational technology • Technology

According to Gagné ( 1974 ) the main question of educational
technology is: How can “things of learning” best be employed
to promote learning? In most discussions of technology
implementation, learning issues remain relatively tacit
(Bransford, Brophy, & Williams, 2000 ) . Searching the rela-
tionship between learning theories and technologies is at first glance an attractive endeavor given its possible relevance for
both educational theory and practice. However, dealing with
this issue is quite complex. Indeed, a number of questions
arise about the relationship of learning theory and technology,
sometimes called a marriage (Perkins, 1991 ; Salomon
& Ben-Zvi, 2006 ) . Do learning theories refer to hybrid con-
structs or are they rather eclectic containers of more modest
models or even common sense practice? How should tech-
nology be conceptualized? If a link exists between learning
and technology, what is the nature of the relationship? Can
we best label developments in the knowledge-base of learn-
ing and technology as paradigm shifts (Koschmann, 1996 ) ,
sequential events (Sloan, 1973 ) , or waves (Tof fl er, 1980 ) ?
In this chapter we will not reiterate broad accounts of evo-
lutions in educational technology (see amongst others De
Corte, Verschaffel, & Lowyck, 1996 ; Januszewski, 1996 ;
Kozma, 1991 ; Mayer, 2010 ; Molenda, 2008 ; Reiser &
Gagné, 1983 ; Saettler, 2004 ) . We start the quest for linking
learning theories and technologies at the moment explicit
learning theory enters educational technology. The critical
appraisal of the link between learning theories and technologies
is structured around the following observations to reduce
complexity and fuzziness in that interdisciplinary fi eld: (1)
evolutions in society and education have in fl uenced the
selection and use of learning theories and technologies; (2)
learning theories and technologies are situated in a somewhat
vague conceptual fi eld; (3) learning theories and tech-
nologies are connected and intertwined by information
processing and knowledge acquisition; (4) educational tech-
nologies shifted learner support from program or instructor
control toward more shared and learner control; and (5)
learning theories and fi ndings represent a fuzzy mixture of
principles and applications.
Observation 1: Evolutions in Society and Education Have In fl uenced the Selection and Use of Learning Theories and Technologies
Educational technology in fl uenced in many and often cen-
trifugal ways educational innovation as part of societal devel-
opment. Successive behaviorist, cognitive, constructivist,
and socio-constructivist approaches to learning and the con-
comitant use of technologies suggest a clear, straightforward
contribution to education based on the internal dynamics of
that fi eld. However, one may wonder why in the 1960s and
1970s behavioral learning theory, but no others, was selected
as the focus of educational technology. Examples of more
cognitively oriented theories available at that time are the
work of Bartlett ( 1958 ) on “Thinking, an experimental and
social study,” of Bruner ( 1961 ) on “The act of discovery,” of
de Groot ( 1965 , originally published in 1946) on “Thought
and choice in chess,” of Dewey ( 1910 ) on “How to think,” of
Piaget ( 1952 ) on “The origins of intelligence in children,”
and of Vygotsky ( 1962 , originally published in 1934) on
“Thought and language.” These theories inspired school cur-
ricula and teaching methods but not technology use. Even
though Newell and Simon ( 1972 ) contend that the appear-
ance of modern computers at the end of World War II gave
researchers the courage to return to complex cognitive per-
formances, there was no relationship between early cognitive
research and technology for education.
It is clear that more than learning science controls the
selection and use of peculiar learning theories and tools. This
points to the impact of society on educational technologies in
that learning theories are selected to support the technology
implementation society drives us to employ (Boyd, 1988 ) .
Indeed, society holds strong expectations to solve learning
problems with technology. Expectations function as macro-
hypotheses that are progressively shaped and falsi fi ed during
implementation, often resulting in more dif fi culties and less
productivity than initially expected. One waits for the next,
more powerful learning theory or tool (Lowyck, 2008 ) .
The in fl uence of the Zeitgeist can be illustrated with some
examples. At fi rst, audiovisual tools were expected to bring
reality into the stuffy classroom and to bridge the gap between
school and the world outside the classroom. Mass media
(radio, fi lm and television) were proclaimed to refresh edu-
cation with real-world information presented just-in-time
(Dale, 1953 ; Saettler, 2004 ) . The audiovisual movement was
grounded on communication theories that model the fl ow of
interaction between sender and receiver, regulating the trans-
port of information (Kozma, 1991 ; Levie & Dickie, 1973 ;
Saettler, 2004 ; Tosti & Ball, 1969 ) . While this movement
nicely illustrates the impact of societal expectations on edu-
cation, no explicit learning theory provided a foundation, so it is not part of our critical appraisal of linking learning
theories and technology.
At the end of the 1950s in the aftermath of the Sputnik-
shock, Western societies aimed at improving education qual-
ity especially in mathematics and science to compensate for
the supposed failure of the progressive education movement
and teachers’ de fi cient classroom behaviors (Skinner, 1968 ) .
In line with the back – to – basics movement (Boyd, 1988 ) , curricula were revised and proper, programmed design and
delivery of subject-matter was expected to contribute to edu-
cational quality based on a genuine science of instruction
(Glaser, 1965 ; Lockee, Larson, Burton, & Moore, 2008 ) . In a similar vein, democratization of education was aimed at
giving increased access to education responding to the
post-war baby boom which led youngsters in a prosperous
economic period to mass education. This, however,
raised concerns about individual development though inter-
preted in multiple ways by Rousseau-inspired romantics to
more mechanistically oriented empirical behaviorists
(Grittner, 1975 ) . Computer-assisted instruction (CAI) claimed to realize individualization which brought Suppes
( 1969 ) to expect that computers could offer individualized
instruction, once possible for only a few members of the aris-
tocracy, to all students at all levels of abilities. However, the
limited capacity of computers and reductionist instructional
design at that time hindered the full implementation of
In the late 1970s, increasing use of personal computers in
professional settings responding to the challenges of an
information society created a new argument for the integra-
tion of computers in education and emphasis on acquiring
computer skills (Dillemans, Lowyck, Van der Perre, Claeys,
& Elen, 1998 ; Mandinach, 2009 ) . This is why policy-makers
in most Western countries launched extensive national pro-
grams to introduce new technologies in schools (Kozma,
2003 ) . Learning to program computers, for example, was seen as a main task for education in a growing technology-rich society. Teachers and other computer savvy practitioners built instructional materials based on common sense knowledge of classroom teaching and content delivery with simple ques-
tion-answer-feedback loops, vaguely inspired by behavioral
principles (Saettler, 2004 ) . This led to a proliferation of small
and isolated CAI-programs, mostly in algorithmic subjectmatter
domains with little theoretical underpinnings or fundamental
goals to achieve (McDonald & Gibbons, 2009 ) .
The interplay between behaviorist learning theory and tech-
nology ultimately resulted in in fl exible and didactic instruc-
tion (Shute & Psotka, 1996 ) .
During the 1980s, a cognitive orientation in education
was strongly supported by Western governments struggling
with increasing worldwide competition in commerce, indus-
try, science and technology. Enhancing learners’ common
understandings of complex issues, deep learning and com-
plex skillfulness instead of mere subject-matter delivery was
perceived as a strategic approach to societal survival (NCEE,
1983 ; Sawyer, 2006 ) . This shift resulted in more complex
forms of cognitive behavior embedded in school curricula,
increasing interest in the role of knowledge in human behav-
ior, and an interactionist view of learning and thinking
(Resnick, 1981 ) . The ambition to tune education by means of
technology to complex changes in society gave birth to a
new wave of investments in research and development not
only in supplying funds and resources for equipment and
network connectivity (Jones, 2003 ) . Many computer micro-
worlds, cognitive tools and instructional programs were pro-
duced at research centers, universities and enterprises (Duffy,
Lowyck, & Jonassen, 1993 ) . However, most of these com-
puter-based educational systems were not widely adopted or
embraced. This was due to both the not – invented – here syn-
drome and the increasing cost of commercial products (Boyd,
1988 ; Jonassen, 1992 ) .
Intensive electronic networking, and social media re fl ect
more recent changes in society that are expected to add value
through a common purpose and deliberate collaborative
action in a community of learners and practitioners (Center
for Technology in Learning, SRI, 1994 ) . Increasing minia-
turization, integrated functionalities, and wireless use com-
prise a communication hyperspace in a global world that call
for new ways of technology use in education. This is why
socio-constructivist theories and technology-supported com-
munities of learning and practice have become dominant, at
least as a frame of reference within the community of educational
Evolutions of learning theories and technologies show inter-
nal and autonomous dynamics that lead toward mutual fer-
tilization. Pressure in Western countries to survive in a
scienti fi cally and economically changing, competitive world
activates governmental initiatives to support technology in
schools through fi nancial support and stimulation of research
and development. However, policy makers often formulate
unrealistic expectations due to lacking knowledge of the
multidimensionality of technological solutions for education.
Commercial organizations respond to societal demands
with little concern about ef fi ciency, effectiveness and rele-
vance of educational products and processes, an observation
that brings researchers to request grounded evaluation
(Clark, 1983 ; Salomon, 2002 ) . Schools and educational
institutions are involved in lasting and dif fi cult processes of
innovation through technologies that impact all organiza-
tion components (curricula, personnel, fi nances, infrastruc-
ture, etc.), while teachers and learners are challenged to
cultivate new competencies, unlearn dysfunctional behav-
iors and conceptualizations, and build new perspectives on
technologies for learning.
Observation 2: Learning Theories and Technologies Are Situated in a Somewhat Vague Conceptual Field
Exploring links between learning theories and technology is
dependent on agreed upon conceptual frameworks and con-
cepts within research traditions. Each fi eld of study is fi lled
with ill-de fi ned concepts and terminology that is inconsis-
tently used and leads toward different starting positions.
A basic science of learning starts from the insight that little is
known and that much has to be discovered, while applied
science and technology focus on what is known and appli-
cable in practice (Glaser, 1962 ) . Despite continuous efforts
to calibrate conceptual issues (Januszewski & Persichitte,
2008 ; Reiser & Ely, 1997 ) , and unlike the natural sciences,
concepts in the behavioral sciences are rarely standardized (Halliday & Hasan, 1985 ) . That concepts are used in various
ways becomes especially problematic when central theoreti-
cal importance is involved (Prenzel & Mandl, 1993 ) .
Learning Theories
Learning as a relatively permanent change in motor, cogni-
tive and psychodynamic behavior that occurs as a direct
result of experience is shared by all learning theories. Despite
this largely accepted de fi nition, “learning theory” remains a
broad term with many perspectives “ranging from funda-
mental exploratory research, to applied research, to techno-
logical development, through the speci fi cation of work-a-day
methods of practice” (Glaser, 1965 , p. 1). Conceptual confu-
sion originates partly from an over-generalization of succes-
sive ways of thought that are perceived as the way things are.
Observable behavior, mind, information processing, socio-
cultural theories, genetics and brain research are changes
that signal scienti fi c progress but the tendency to over-gener-
alize is often driven by other than scienti fi c considerations
(Bredo, 2006 ) . Given the intrinsic limitations of educational
research, no single theory encompasses all aspects of learning
and learners (Gage, 1972 ) . Consequently, various theo-
ries that emerged as researchers focused on different kinds of
learning only represent a limited part of the knowledge-base
of psychology as a discipline (Bransford et al., 2006 ) . In
addition, learning theories do not constitute a monolithic,
coherent system but each school of thought represents a col-
lection of distinct theories that are loosely connected (Burton,
Moore, & Magliaro, 1996 ; Dede, 2008 ) a fact that led to the
balkanization into smaller communities with different
research traditions and largely incommensurable views of
learning (Koschmann, 1996 ) . While behavioral theory and
early information processing theory use de fi nitions that are
instrumental to experimental research, socio-constructivist
theory is complex, eclectic, and multifaceted (Lowyck &
Elen, 1993 ) . A possible solution is to take a pragmatic posi-
tion de fi ning learning theories as an interrelated set of facts,
propositions, rules, and principles that has been shown to be
reliable in many situations (Spector, 2008 ) . Though this may
be helpful to avoid conceptual fuzziness, it seems hard to
de fi ne valid and precise criteria to differentiate between evi-
dence-based and common sense knowledge in an educational
Educational technology holds a double meaning: (a) applica-
tion of scienti fi c know-how, and (b) tools or equipment
(Glaser, 1965 ; Molenda, 2008 ; Reiser & Gagné, 1983 ) .
AECT (the Association for Educational Communications
and Technology; http://www/ ) refers to the
“disciplined application of scienti fi c principles and theoreti-
cal knowledge to enhance human learning and performance”
(Spector et al., 2008 , p. 820), which is very close to instruc-
tional design as de fi ned by Gagné ( 1974 ) as a “body of tech-
nical knowledge about the systematic design and conduct of
education, based upon scienti fi c research” (p. 3). Technology
as the mere application of research fi ndings was highlighted
in the years of programmed instruction with procedures for
behavioral modi fi cation to reach terminal behaviors (Glaser,
1965 ) . Along with an increasing variety of learning theories,
different genres of technology-based learning environments
covered different functions of educational technology,
including intelligent tutoring systems, interactive simula-
tions and games, animated pedagogical agents, virtual envi-
ronments, and computer-supported collaborative learning
systems (Mayer, 2010 ) .
Others focus on the physical aspects of technology via
which instruction is presented to the learners. McDonald and
Gibbons ( 2009 ) refer to this as the tools approach which
holds the expectation that using technological tools will
affect learning outcomes. This led to various gimmicks being
introduced in schools as extras not necessarily well aligned
with the teaching-learning process (Husèn, 1967 ) . Machines
on their own will not bring about any change (Stolurow &
Davis, 1965 ) . This statement is close to Clark’s ( 1983 ) view
that method, not media, determines effectiveness. This claim
also pertains to the comparison of computer-based environ-
ments (e.g., desktop simulation and virtual reality simulation)
(Mayer, 2010 ) . The question, however, is not if tools
can contribute to learning but how instructional materials in
various forms can enhance learning and allow the manipula-
tion of the properties of instruction that impact learning
(Lumsdaine, 1963 ) . This re fl ects the position of Kozma
( 2000 ) who emphasizes a nexus of media and method.
Indeed, technology allows for methods that would not other-
wise be possible, such as interactive multimedia simulations
that support the ability to act on the environment and not
simply observe it (Winn, 2002 ) or hypermedia that challenge
cognitive fl exibility while crisscrossing the information land-
scape (Spiro, Feltovich, Jacobson, & Coulson, 1992 ) . In
times when information and communication technologies
deeply penetrate society, the dichotomy between applied sci-
ence and tools technology has been in favor of synergy.
Educational technology involves a broad variety of modali-
ties, tools, and strategies for learning (Ross, Morrisson, &
Lowther, 2010 ) .
Linking Learning Theories and Technology
Given the complexity and diversity of conceptualization,
it seems dif fi cult to fi nd a direct link between learning theories and technology. Firstly, the relationship is asymmet-
ric; it is common to consider learning theories as leading and
technology as following (Salomon & Perkins, 1996 ) . Secondly, although the psychology of learning is a critical
foundation area (Spector, 2008 ) , in complex technological
environments it shares a place with communication theory,
general systems theory, and instructional-curriculum theory
(Richey, 1986 ) . In fact, not only learning but organizational
issues as well are important in technological environments,
with a focus on the availability, accessibility and acceptabil-
ity of educational resources (Lane, 2008 ) . An analysis of
articles in journals on educational psychology between 2003
and 2007 shows that only 5.6 % of the articles addressed the
links between learning theory and technology (Nolen, 2009,
as cited in Ross et al., 2010 ) .
Learning theories, technology and their interlinking fi elds
are dependent on speci fi c research traditions, historical arti-
facts, idiosyncratic frameworks, technology-based func-
tionalities and pragmatics, which necessarily leads to
divergence. Calibration of concepts and conceptual frame-
works is not merely a philosophical issue but it is critical for
cumulative knowledge building. Not surprisingly, rapid
changes in learning theories and technologies generate new
terminology. However, increased efforts to re fi ne concep-
tual frameworks for valid theory building are needed to sup-
port cumulative domain knowledge in the fi eld of educational
technology. Given the conceptual complexity, the expecta-
tion that a clear link between learning theories and
technology can be built based on agreed upon de fi nitions is
in vain. Consequently, a solution has to be found in a more
pragmatic approach with a smaller unit of analysis, where
(partial) learning theories, models, and principles are con-
nected to speci fi c technological tools in order to overcome
conceptual overload.
Observation 3: Learning Theories and Technologies Are Connected and Intertwined by Information Processing and Knowledge Acquisition
Different learning theories and epistemologies (e.g., objec-
tivism and constructivism), lead to various conceptions of
information processing and knowledge acquisition that
in fl uence technology use. Given the central function of edu-
cation to help learners acquire declarative, procedural and
conditional knowledge, learning theories and technologies
are fellow travelers.
Behaviorist Theory and Subject-Matter Decomposition
In the behaviorist tradition, knowing is an accumulation of
associations and components of skills that prescribes simpler
tasks as prerequisites for more complex ones (Greeno,
Collins, & Resnick, 1996 ) . The stimulus-response theory in
which knowledge is de fi ned as a learner’s collection of
speci fi c responses to stimuli that are represented in behav-
ioral objectives is basic in programmed instruction and CAI.
Logical presentation of content, requirement of overt
responses, and presentation of immediate knowledge of cor-
rectness are common characteristics. Subject-matter is
decomposed into small units with carefully arranged
sequences aimed at speci fi ed terminal behaviors (Shrock,
1995 ) . Terminal behaviors are de fi ned as understanding con-
cept formation, concept utilization, and reasoning through
variations of the stimulus context (Glaser, 1962 ) , not through
direct access to thinking or knowledge organization.
Researchers and designers massively invested in re fi ning and
shaping the initial principles of content framing and sequencing
(Lockee, Moore, & Burton, 2004 ; Tennyson, 2010 ) .
A frequently cited example of a system based on behaviorist learning theory, is programmed logic for automatic teaching
operation (PLATO), a mainframe-based, integrated system
of hardware and software with well-designed instructional
materials displayed on special terminals connected through
satellite links. The PLATO system started in the early 1960s;
didactic as well as communication functions were gradually
expanded (Molenda, 2008 ; Saettler, 2004 ) , leading to over
15,000 h of instructional materials available in a variety of
disciplines (Simons & de Laat, 2006 ) . Despite its continu-
ous adaptation and extension, PLATO as a closed system
had to compete with a steady innovation of subject-matter
and didactic approaches in curricula, a paradigm shift toward
a cognitive interpretation of learning environments, and a
knowledge-building epistemology. Besides evolutions in
education, fi nancial issues played an important role since
CAI was signi fi cantly more expensive than conventional
instruction and no return on investment was realized
(Saettler, 2004 ) .
Information Processing Theory and Problem-Solving Tasks
Gagné ( 1974 ) dates the transition from behaviorist learning
toward cognitive theory at the moment learning is conceived
of as a matter of students’ information processing. Cognitive
theory is largely rooted in objectivist epistemology, but
unlike the behaviorists, cognitive psychologists emphasize
the individual’s processing of information and how knowledge is stored and retrieved (Winn, 2004 ) . A human
information processing system consists of a sensory register,
short-term (working) memory and long-term-memory
(Simon, 1978 ) . Information moves through stages in the cog-
nitive system with processes and mental representations that
operate at each step (Brown, 1978 ; Glaser, 1991 ) . Mental
processes mediate what is selected, processed, remembered,
recalled and generalized (Hanna fi n & Hill, 2008 ) .
Theory on information processing and problem solving emerged with the development of the digital computer after World War II (Newell & Simon, 1972 ; Simon, 1978 ) and is strongly related to content: “If content is a substantial determinant
of human behavior—if in fact the message is a lot more message than the medium—then information processing theories
have opportunities for describing human behavior veridi-
cally that are foreclosed to theories unable to cope with content
(Newell & Simon, 1972 , p. 11). All problem-solving behavior is framed by the information-processing system, the task envi-
ronments and the problem space (Simon, 1978 ) . In a cognitive perspective, knowledge that supports
understanding differs from information as disconnected
facts and formulas (Bransford et al., 2000 ) . There is a clear
shift from information delivery toward student’s knowledge
activation since the logical type of knowledge that was asso-
ciated with a given discipline in a behaviorist approach is
replaced by the psychological nature of meaningful knowl-
edge held by learners (Shuell, 1992 ) . Subject-matter is no
more fragmented in small parts but organized around prob-
lems that activate learner’s prior declarative, procedural,
and self-regulatory knowledge in an interconnected way to
solve a given problem. Processing and transformation capa-
bilities of computer micro-worlds allow learners to progress
unto more advanced models, increasing the number of rules,
quali fi ers, constraints to be taken into account, and the range
of problems that can be accommodated (Kozma, 1991 ; Seel,
2006 ) . Computer simulations are compatible with a cogni-
tive theory of learning since they present formalized mod-
els, elicit speci fi c cognitive processes like hypothesis
generation and testing, allow for learner activity in terms of
model manipulation, and interact with the underlying
domain model. Learners can execute actions like changing
the values of input variables, observing the effects in output
variables and make or test hypotheses based on the changes
in values that foster conceptual change (de Jong, 1991 ;
Winn, 2004 ) .
Cognitive Theory and Knowledge Organization
Knowledge is a complex phenomenon involving such con-
structs as schema, mental models, symbol manipulation,
knowledge construction, and conceptual change (Winn,
2004 ) . Research in cognitive psychology revealed the
centrality of knowledge in human performance including
content, knowledge structure and context (Cooke, 1999 ) .
Knowledge in isolation (inert knowledge) is of little value
but knowledge is powerful if highly organized and easily
accessible (Greeno et al., 1996 ; Schraw, 2006 ) . However,
reduction of knowledge organization to neat hierarchies and
sequences is an oversimpli fi cation of the knowledge people
construct (Siemens, 2004 ; Winn, 1993 ) . Indeed, each indi-
vidual must possess extensive knowledge, organize knowl-
edge into interconnected schemata and scripts, and use that
knowledge to construct conceptual mental models of a given
subject-matter domain that are used to solve problems and
think critically (Schraw, 2006 ) .
Knowledge organization has been supported by different
cognitive tools, such as simulations (de Jong, 2010 ) , concept
mapping, and semantic networking embedded in computer
tools that visually represent a cognitive structure with nodes
and links (Jonassen & Reeves, 1996 ) . In the early 1990s, sev-
eral computer-based tools were developed (Kommers,
Jonassen, & Mayes, 1992 ) challenging learners to analyze
structural relationships among the subject-matter. “Learning
tool” (Kozma, 1992 ) , “TextVision” (Kommers & de Vries,
1992 ) , and “SemNet” (Fisher, 1992 ) are examples of soft-
ware packages that allow users to graphically represent con-
cepts, de fi ne relationships and enter detailed textual and
graphic information for each concept. However, a graphical
representation of knowledge structure is limited both in mir-
roring knowledge complexity and accessing deep knowl-
edge. The complexity of digging up and representing
concepts, nodes and knowledge structures not only accounts
for novices with limited domain knowledge but also for
experts as has been evidenced by research on expert knowl-
edge acquisition (Cooke, 1999 ) .
Constructivist Theory and Knowledge Construction
Knowledge construction is a generative learning process
(Wittrock, 1974 ) . From a constructivist perspective, knowl-
edge is not conceptualized as a body of information based on
veri fi ed facts but, rather, as individually constructed by
observation and experimentation. Knowledge acquisition is
dynamic rather than static, multidimensional rather than
linear, and systemic rather than systematic (Winn, 1993 ) .
The active interaction between an individual and the envi-
ronment is mediated through cognitive structures of the indi-
vidual (Jonassen, Mayes, & McAleese, 1993 ) . The knowledge
that each student constructs is not predictable from the individual
pieces of information in the information landscape or
the curriculum but emerges from the sum of the encounters
and from the relations established by the student within the
knowledge domain. If the learner is seeking information to solve a problem or
build a better understanding, then environments, such as
hypertext retrieval systems, can support that need and engage
the learner. Information retrieval is supported by the learner’s
ability to follow a particular path and make decisions about
which links to follow within the hypertext information. In
order to make learners able to amend the information in some
way, many hypertext systems include functions to support
the creation or editing of nodes and links and other functionalities
(Jonassen, 1992 ) . Learning from hypertext mostly is
task driven, in contrast with free browsing. This is why cog-
nitive fl exibility that allows crisscrossing the information
landscape is not well suited for novices in a given subject-
matter domain (Spiro et al., 1992 ) . Browsing in a domain for
which no properly developed schemata have yet been con-
structed by the learner is not likely to lead to satisfactory
knowledge acquisition at all (Jonassen et al., 1993 ) .
Socio-constructivist Theory and Distributed Knowledge
The information-processing approach with cognition mainly
conceived of as involving internal mental processes came
under increasing criticism. The main objection was that
knowledge can be viewed as distributed over individuals and
their environments rather than as something self-suf fi cient to
an individual. The notions of distributed cognition and distributed knowledge play an important role as human activity
is affected by contextual affordances which include both
people and cultural artifacts (Greeno et al., 1996 ; Hewitt &
Scardamalia, 1998 ; Säljö, Eklund, & Mäkitalo, 2006 ) . Glaser
( 1991 ) offers several arguments for integrating the social
dimension within a cognitive perspective: (a) available
knowledge is extended; (b) the loci of self-regulatory activity
are multiplied; (c) learners can help each other in realizing a
Vygotskian zone of proximal development; and (d) a social
context helps in bringing thinking to an observable status.
The socio-constructivist perspective and the distributed
character of knowledge have in fl uenced computer use since
about 1990. CSCL (computer supported collaborative learn-
ing) serves groups of learners who co-construct knowledge
in a given subject-matter context and aim at goals that are
externally provided. CSCL technology is used to present or
stimulate a problem for study, helping to situate it in a real-
world context, mediate communication within and across
classrooms, provide archival storage for the products of
group work, or enable learners to model their shared under-
standing of new concepts (Koschmann, 1996 ) .
Computer-supported intentional learning environment
(CSILE) and its extension Knowledge Forum are instances
of CSCL that encourage structured collaborative knowledge-
building instead of focusing on individual learning tasks
(Scardamalia & Bereiter, 1994 ) . Students communicate ideas
and re fl ections, ask questions, exchange statements and con-
tinuously build up shared knowledge as input in a database.
The computer system supports the knowledge organization
of individual and community discourse. The target is real
world knowledge that is constructed over time and not
restricted to a single product or topic (Scardamalia, 2002 ;
Siemens, 2004 ) .
Conceptions of information processing and knowledge build-
ing change over time, depending on epistemological argu-
ments and evolving learning theories. Different computer
tools and systems have been designed to contribute to the
supposed increase of education quality in terms of knowledge
acquisition but most if not all are limited in curriculum cov-
erage. The shift from programmed instructional materials as
parts of the school curriculum toward student’s individual
and collective knowledge organization and knowledge con-
struction tools paved the way for more real-world problems
and knowledge. Evaluation studies clearly show that not only
the use of cognitive tools but the link with underlying cogni-
tive processes de fi nes a system’s or a tool’s merits.
Observation 4: Educational Technologies Have Shifted Learner Support from Program or Instructor Control Toward More Shared and Learner Control
A basic tenet in the discussion of the interplay between technology and education is how technology might support individuals and groups to reach learning goals. Depending upon available learning theories and technological tools, different kinds of support have been inserted into instruc-
tional materials, programs, and technology enhanced learning
environments, while open-ended learning environments suggest freedom to learn . This reveals a tension between structured learning support and a learner’s self-management with technology.
Intelligent Computer-Assisted Learning and Intelligent Tutoring Systems
In the behaviorist tradition, computers integrate the activities
of a display component, a response component and a feedback
component of instruction (Gagné, 1974 ) . It was expected
that computer-assisted learning could realize maximal learn-
ing support through adaptive feedback. However, linear
feedback often results in deficient individual support in traditional CAI programs. A solution is sought in the design
of a new generation of programs called intelligent computer-
assisted instruction (ICAI). They are instances of micro-
adaptive instruction that aim at continuously tuning
instruction to the needs of the individual learner with branch-
ing as a fundamental aspect of design (Wenger, 1987 ) . ICAI
systems are behavioristic since they only use the status of
student’s behavior to adapt instruction (Urban-Lurain, 1996 ) .
However, genuine feedback is hard to realize since the source
of information is external to the student and takes place not
during a learning activity but only after task completion
(Butler & Winne, 1995 ) . In addition, limited computer capac-
ity in terms of memory and speed imposed severe restrictions
to tune feedback to individual needs of students.
Fine-tuned adaptivity based on a student’s cognitive status had to wait for intelligent tutoring systems (ITSs). A cognitively oriented tutoring system or ITS is not a static
preprogrammed system but integrates computational models
using arti fi cial intelligence and cognitive science to generate
interventions. These are generated based on data gathered
from a database that includes the nature of errors and cognitive
skills that are realized in the form of production rules
(Shute & Psotka, 1996 ) . The database is structured around
(a) an expert or domain model, (b) a dynamic student model,
(c) a tutor or teaching model, and (d) a communication model
and user interface (De Corte et al., 1996 ; Larkin, 1991 ) .
Anderson ( 1983 ) developed his adaptive control of thought
(ACT*) theory in which a learner’s knowledge is tracked
(knowledge tracing) in order to generate appropriate learn-
ing activities.
In ITSs two different lines of evolution can be observed.
One is to re fi ne ITSs in order to integrate new knowledge
about learning and new programming techniques. The other
is the acceptance of limitations since intelligent machines do
not have the breadth of knowledge that permits human rea-
soning given the fuzziness of thinking and permeability of
the boundaries among cognitive schemata (Winn, 2004 ) .
Progress has been made in ITS development mainly in
knowledge-domains with a rule-based, logical structure,
such as classical mechanics, geometric optics, economics,
elementary algebra, grammar, and computer programming
(Sleeman & Brown, 1982 ; Wenger, 1987 ). Further development
of natural language processing (Graesser, Chipman, & King,
2008 ) allows the ITSs to make decisions based on qualitative
data analysis (e.g., open-ended text responses or annotated
concept maps) (Lee & Park, 2008 ) . Implementations of such
ITSs are found in (a) adaptive hypermedia systems (AHSs)
which combine adaptive instructional systems and hyperme-
dia-based systems (Brusilovsky, 2001 ; Lee & Park, 2008 ;
Vandewaetere, 2011 ) , (b) affective arti fi cial intelligence in
education (AIED) to detect and intelligently manage the
affective dimension of the learner (Blanchard, Volfson,
Hong, & Lajoie, 2009 ) , (c) Web-based AHSs that adapt to
the goals, interests, and knowledge of individual users
(Brusilovsky, 2007 ) , (d) intelligent simulation learning envi-
ronments with advanced help, hints, explanations and tutor-
ing facilities (de Jong, 1991 ) , and (e) sophisticated online
courses that incorporate intelligent tutoring systems
(Larreamendy-Joerns & Leinhardt, 2006 ) .
Notwithstanding large investments and re fi ned adaptivity, the ITS movement was in decline. Firstly, ITSs can model pro-
cedural skill acquisition but they show limitations in simulating student’s complex cognitive processes and situated activity. Secondly, computer-based tutoring systems resulted in many highly structured, directive systems due to the limitations of ITSs to simulate ill-structured or not-rule-based domains (Shute
& Psotka, 1996 ) . The consequence is that if computer simula-
tion is impossible, then so is intelligent tutoring. This led
Kintsch ( 1991 ) to launch the idea of “unintelligent” tutoring in which a tutor should not do all the planning and monitoring
because these are activities that students must perform in order to learn. In this view, computers tools, though not arti fi cially intelligent, can play a role to support mindful processes in students
(Derry & Lajoie, 1993 ; Jonassen, 2003 ; Jonassen & Reeves, 1996 ; Salomon, Perkins, & Globerson, 1991 ) .
Computer-Enhanced Learning Environments and Learner Support
Transition from instructional materials or programs to learn-
ing environments brings about a shift in the locus of control
from system to learner which in fl uences the role of system
intelligence to support the learner (Chung & Reigeluth, 1992 ;
van Joolingen, 1999 ) . Locus of control can be classi fi ed as
external (program control), internal (learner control) or
shared (Corbalan, Kester, & van Merriënboer, 2008 ; Elen,
1995 ; Hanna fi n, 1984 ; Lawless & Brown, 1997 ) . In contrast
to ITSs as a mode of program-based guidance, learning envi-
ronments allow learners to reify a learning process while
maintaining task complexity (Bereiter & Scardamalia, 2006 ;
Collins, 1996 ; Zucchermaglio, 1993 ) . Learner control allows
learners to make instructional decisions on support needed
and content to be covered, choosing the estimated optimal
level of dif fi culty, sequencing a learning path, regulating
both the kind and speed of presentation, and de fi ning the
amount of information they want to process (Dalgarno, 2001 ;
Merrill, 1984 ; Vandewaetere, 2011 ) .
Multiple descriptions of constructivism suggest divergent
ways to interpret and operationalize learner support.
Discovery learning, problem-based learning, inquiry learning,
experiential learning and constructivist learning are versions
of open learning that leads to the perception that almost
unlimited control can be given to students (Bednar, Cunningham, Duffy, & Perry, 1991 ; Honebein, Duffy, &
Fishman, 1993 ; Kirschner, Sweller, & Clark, 2006 ) . This view is rooted in the work of radical constructivists such as
Papert ( 1980 ) who points to the paradox that new technolo-
gies, instead of creating opportunities for the exercise of
qualitative thinking, tend to reinforce educational methods
whose very existence re fl ect the limitation of the pre-com-
puter period. In his view, based on his collaboration with
Piaget, learning as self-discovery with Logo as a tool can
occur without being taught. His strong constructivist posi-
tion holds that “In the Logo environment … the child is in
control: The child programs the computer. And in teaching
the computer how to think, children embark on an explora-
tion about how they themselves think” (p. 19). In his opin-
ion, the acquisition and transfer of programming skills
induced by Logo would happen to the pupils (De Corte,
Verschaffel, Schrooten, & Olivié, 1993 ) . Studies on that cog-
nitive-effects hypothesis of Logo on children did not deliver
positive results (De Corte, 1996 ) . Most researchers share the
viewpoint that systematic guidance and even direct instruc-
tion needs to be embedded in the program with ample room
for exploration. In his reaction to the fi ndings, Papert ( 1987 )
ascribes the criticism that Logo did not deliver what it prom-
ised to a technocentrist, rigourous model of research: “The
fi nding as stated has no force whatsoever if you see Logo not
as a treatment but as a cultural element—something that can
be powerful when it is integrated into a culture but is simply
isolated technical knowledge when it is not” (p. 24). This
illustrates the lasting problem with constructivism and all its
derivatives as an ideology as opposed to a learning theory.
Even in a constructivist framework, students have goals to
pursue (Clark, Kirschner, & Sweller, 2012 ; Winn, 1993 ) , be
they externally or internally generated.
More moderate conceptions of control can be found with
learners as partners in distributed intelligence to enhance
cognitive and metacognitive knowledge and strategies
(Salomon et al., 1991 ) . Examples of constructivist learning
environments with explicit learner support are cognitive
apprenticeship and situated cognition (Collins, Brown, &
Newman, 1989 ) , anchored instruction (Cognition and
Technology Group at Vanderbilt, 1993 ) , and simulation
learning environments (de Jong, 1991 ) . They contain
advanced help, hints, modeling, coaching, fading, articula-
tion, re fl ection, and exploration to support the process of
increasing learner control. In order to counter helplessness in
multimedia, standard pop-up help systems, animated guides
or intelligent agents that monitor browsing patterns of learn-
ers are designed (Dalgarno, 2001 ) .
Learner support has been realized in different computer-
based learning contexts from which two are exempli fi ed: (a)
use of computer tools that originated outside education (De
Corte et al., 1996 ; Duffy et al., 1993 ) , and (b) dedicated tools
embedded in the environment (e.g., pedagogical agents)
(Clarebout, Elen, Johnson, & Shaw, 2002 ) . Publicly available
computer tools have been inserted into many learning
environments (e.g., word processors, calculators, spread-
sheets, database programs, drawing and composition pro-
grams) to free students from the intellectual burden of
lower-level operations, present a familiar structure for per-
forming a process, and trace states and processes so as to
contribute to the quality of a student’s thinking and learning
(Jonassen, 1992 ) . The supply of tools has been enlarged with
WebQuests, simulations and games, micro-worlds, blogs,
and wikis (Molenda, 2008 ) , and social media (Säljö, 2010 )
that allow for high levels of interactivity, interactive data
processing, symbol transformation, graphic rendering, infor-
mation storage and retrieval, and communication (Dalgarno,
2001 ; Kozma, 2000 ; Mayer, 2010 ) .
Animated pedagogical agents illustrate endeavors to
embed learner support in interactive learning environments
to enable the system to engage and motivate students by
adapting support to individual students and providing stu-
dents with nonverbal feedback (Johnson, Rickel, & Lester,
2000 ) . Functionalities of learning support delivered by ani-
mated pedagogical agents include supplanting, scaffolding,
demonstrating, modeling, coaching and testing, but meta-
cognitive support is lacking (Clarebout et al., 2002 ) . A pos-
sible explanation for the absence of metacognitive support is
that the design of pedagogical agents stems from the ITS
tradition with a strong focus on domain speci fi c knowledge
and single solution procedural tasks (Clarebout et al., 2002 ) .
Open-Ended Computer Environments: Conditions to Be Met by Learners
Advances in computer technology and multimedia allow
learning experiences with authentic, real-world problems in
which learners have control over activities, tools and
resources (Reiser, 2001 ) . When constructivism is considered
to be a learning theory, most authors interpret it as individu-
als who have to create their own new understandings
(Resnick, 1989 ) though this does not necessarily imply
unguided or minimally guided learning (Mayer, 2004 ; Winn,
1993 ) . Learning environments are goal oriented, which
makes learner’s self-regulation and external support crucially
dependent upon a student’s ability. Student use of support in
open learning environments is not an objective nor an exter-
nal measure, but it is mediated by many characteristics and
processes such as prior knowledge of subject matter, self-
regulating capacity and perspectives on learning environ-
ments and support (Elen & Lowyck, 1998 ; Lowyck & Elen,
1994 ) . High achievers who are knowledgeable about a sub-
ject-matter area can bene fi t from a high degree of learner
control whereas learners who lack knowledge about the
structure of the domain and metacognitive knowledge and
strategies make poor choices (Collins, 1996 ) . Initial schema
development and knowledge acquisition normally must be guided more than advanced knowledge acquisition since a
domain for which no properly developed schemata have yet
been constructed is not likely to lead to satisfactory knowl-
edge acquisition at all (Jonassen et al., 1993 ) . Freedom of
movement in hypermedia can cause inexperienced learners
to get lost in hyperspace ( Spiro et al. 1992 ) . Functionalities
of learning environments, including learner support, seem
effective when learners are in tune with the intentions of the
system and make use of available support (Winne, 2004 ) .
Students do not react to objective or nominal stimuli but to
transformed, interpreted stimuli which commonly leads to a
suboptimal use of instructional interventions (Lowyck,
Lehtinen, & Elen, 2004 ) . Students’ perspectives on learning
environments and their epistemological beliefs (Bromme,
Pieschl, & Stahl, 2010 ) may affect outcomes. Gerjets and
Hesse ( 2004 ) hypothesize that a multiplicity of factors
besides the attributes of the learning environment may play a
role (e.g., knowledge prerequisites, learning styles, learner
preferences, motivational orientations, attitudes, epistemo-
logical beliefs, and instructional conceptions). This emphasizes
the role of student’s perspectives, perceptions and
instructional cognition that mediate between a designed
computer-enhanced environment and student’s use of it.
Learner support in technology rich environments is crucial
for learning. Depending upon learning theories and available
technologies, different kinds of scaffolds have been designed.
CAI only used linear sequences, a limitation that has been
overcome in ICAI and ITSs. The advent of cognitive and
socio-constructivist approaches shifted the focus from pro-
gram control to learner and shared control. The complexity of
theoretical frameworks and operational interventions results
in many different support tools. The expectation that open-
ended learning environments in and of themselves would
result in learning is questionable. The zone of proximal devel-
opment concept needs to be considered. A technological
learning environment is not effective by itself; it has to be
adopted by learners in line with their ability, self-management
and perspectives on technological learning environments.
Observation 5: Learning Theories and Findings Represent a Fuzzy Mixture of Principles and Applications
The proposition that a science of learning is fundamental to educational technology has been broadly accepted but it is unclear how bridging both fi elds can be realized. There are, however, arguments to assert that a direct transfer of theory into practice can no longer be expected. Firstly, the nature of learning sciences and instructional technology re fl ects two separate endeavors with different conceptual frame-
works, methods and goals, often labeled as fundamental versus applied which brings Glaser to contend that “the progress of basic science does not insure systematic and
fruitful interplay between basic knowledge, applied research, and subsequent technology” (Glaser, 1962 , p. 3). Learning theories build a descriptive knowledge base while educational technology needs theoretically valid prescrip-
tions to optimize learning (Elen, 1995 ) . Secondly, building a uni fi ed base of knowledge about learning seems unrealis-
tic since successive learning theories show noncumulative characteristics (Elen & Clarebout, 2008 ) and new technolo-
gies have a tendency to get disconnected from fi ndings obtained with older technologies (Hanna fi n & Young, 2008 ) . Though learning theories as an emerging set of notions rather than as a set of empirical fi ndings and micro-
theories can help us to understand complex systems (Calfee, 1981 ) , they are mostly used as a source of veri fi ed instruc-
tional strategies, tactics and techniques. Behaviorism, for example, is grounded in experimental psychology that delivers laboratory fi ndings, and early information process-
ing theory is based on rich data about individual problem solving, both with high internal validity. Constructivism and socio-constructivism fi nd their origins in externally valid ecological settings that re fl ect multiple perspectives, which renders theories complex, multifaceted and diver-
gent. The former theories (behaviorism and cognitivism) resemble rivers fl owing in a riverbed while the latter (con-
structivism and socio-constructivism) resemble a river delta spreading out into many channels.
Learning Theories, Findings, and Principles
Theories supply fi ndings that are the starting point for applied
research and the development of instructional principles and
devices (Ertmer & Newby, 1993 ; Glaser, 1962 ) . A principle
or basic method re fl ects a relationship that is always true
under appropriate conditions regardless of program or prac-
tice prescribed by a given theory or model (Merrill, 2002 ) .
A principle makes a statement about the outcomes instruc-
tion aims at, the conditions required, and the methods that
can be used (Winn, 1993 ) . Evolution of learning theories,
fi ndings, and principles re fl ect different transitions from the-
ory into practice, ranging from convergent to divergent.
Behaviorist Learning Theories, Findings, and Principles
Behavioral theory focuses on basic laws of behavior
modi fi cation. From experimental behaviorist learning theory
it was expected that principles based on the analysis of sim-
ple performances tested in laboratory conditions could be extrapolated to complex forms of learning (Glaser & Bassok,
1989 ) . Skinnerian operant or instrumental conditioning
based on the relationship between stimuli that precede a
response (antecedents), stimuli that follow a response (con-
sequences) and the response (operant) itself has been broadly
accepted in instructional technology (Winn, 2004 ) .
Reinforcement, contiguity and repetition are pivotal in the
acquisition of behavior (Burton et al., 1996 ) which can easily
be translated into behavioral control principles. These prin-
ciples led to agreed upon speci fi cations for instructional
materials like analysis of terminal behaviors, content, objec-
tives, criteria-referenced assessment, learner and behavior
characteristics, sequencing of content from simple to complex, and frame composition (Andrews & Goodson, 1980 ;
Ertmer & Newby, 1993 ; Lockee et al., 2004 ; Montague &
Wulfeck, 1986 , Tennyson, 2010 , Winn, 1993 ) . Programmed
instruction and CAI are organized in small, easy steps to let
the learner start from an initial skill level and gradually mas-
ter a task while reducing prompting cues along the path to
mastery. More evidence has been collected on the prompting
aspect rather than the fading aspect (Lumsdaine, 1963 ) .
Despite intensive and lasting efforts to implement behav-
ioral principles in instructional environments, the narrow
focus on links between stimulus and response led to a reductionist
and fragmented perspective. However, criticism
should not only be directed at the behavioral foundation but
also at the poorly developed software (Cooper, 1993 ) .
Cognitivist Learning Theories, Findings, and Principles
The invalid expectancy that stimulus-response can account
for complex human behavior (Tennyson, 2010 ; Winn, 2004 )
challenged cognitive learning theory to open the black box
of mental activities (Glaser, 1991 ) . Stimulus-response as the
unit of behavior is replaced by a cognitive interpretation
with emphasis on planning and hierarchical organization of
the mind. Early cognitive learning theories focus on prob-
lem-solving and information processing based on Miller’s
work on chunking and the limited capacity of working
memory (Miller, 1956 ) and the TOTE unit “test-operate-
test-exit” (Miller, Galanter, & Pribram, 1960 ) . Though prob-
lem-solving and information processing are interconnected
fi elds (Newell & Simon, 1972 ) , fi ndings are translated into
separate principles for problem-solving and information
Problem-solving theory was initially elaborated for pro-
cesses of relatively well-structured puzzle-like problems in
laboratory settings in which a given state, a goal state and
allowable operators are clearly speci fi ed (Simon, 1978 ) . This
led to the following principled sequence: (a) input translation
that produces a mental representation, (b) selection of a par-
ticular problem-solving method, (c) application of the
selected method, (d) termination of the method execution,
and (e) introduction of new problems (Newell & Simon,
1972 ) . Studies on complex problem solving revealed some
core instructional principles, such as (a) develop skills within
speci fi c domains rather than as general heuristics (domain-
speci fi c), (b) restrict problem-solving skills to a limited range
of applicability (near-transfer principle), and (c) integrate
different kinds of knowledge within guided problem-solving
tasks (integration principle) (Mayer & Wittrock, 2006 ) .
These principles can be used in designing micro-worlds or
simulations but they hold no indication how to link principles
to tools. Translation of fi ndings into principles and instruc-
tional technology is highly dependent on an instructional
designer’s decisions and available technologies.
Information processing systems describe how people per-
ceive, store, integrate, retrieve, and use information. Findings
from information processing theory mirror principles for
educational technology. They focus on the load that perform-
ing a task causes to a learner’s cognitive system (Mayer,
2010 ; Paas & van Merriënboer, 1994 ; van Merriënboer &
Sweller, 2005 ) . Cognitive load theory is based on assump-
tions about dual-coding (Paivio, 1986 ) , limited working
memory and chunking (Miller, 1956 ) , and cognitive process-
ing for meaningful learning (Mayer & Moreno, 2003 ) .
Examples of such principles are as follows: (a) if the visual
channel is overloaded, move some essential processing from
the visual to the auditory channel; (b) if both visual and auditory
channels are overloaded, use segmenting and pre-training;
(c) if one or two channels’ overload is caused by extraneous
material, use weeding and signaling, and if caused by confusing
presentations, align and eliminate redundancy; (d) if
one or both channels are overloaded by representational
holding, synchronizing and individualizing are useful (Mayer
& Moreno, 2003 ) . These principles are close to the informa-
tion processing theory and can be empirically tested (van
Merriënboer & Sweller, 2005 ) .
The cognitive orientation effectuated a shift from materi-
als to be presented in an instructional system to students’
goal-oriented and self-regulated processes and dialogue with
the instructional design system (Cooper, 1993 ; Merrill,
Kowalis, & Wilson, 1981 ; Merrill, Li, & Jones, 1990 ;
Tennyson, 1992 ) . This shift leads to more general principles
to build cognitive learning environments, like activation of
learner’s involvement in the learning process through learner
control, self-monitoring, revising techniques, cognitive task
analysis procedures, use of cognitive strategies, and allowing
students to link prior and new knowledge (Ertmer & Newby,
1993 ) . In addition, theories and concomitant principles are
dependent on evolutions in technology. While, for example,
early attempts to implement cognitively oriented instruction
in technology tools were inappropriate or ineffective,
increased hardware speed and capacity allowed us to imple-
ment cognitive-based learning using hypertext, hypermedia,
expert systems, and so on (Cooper, 1993 ) . (Socio-) constructivist Learning Theories, Findings, and Principles
Information processing adapts an objectivist epistemology and represents a mechanistic view of learning with ready recall of information and smooth execution of procedures (Perkins, 1991 ) . Increasing complexity and situatedness of learning led to dissatisfaction with the computational view of cognition and the restriction of learning to internal mental representations. This leads to a constructivist per-
spective on learning as the creation of meaning based on
experience-in-context (Bednar et al., 1991 ; Duffy et al.,
1993 ) . Constructivism as an umbrella term holds many per-
spectives and approaches, including situated cognition, realis-
tic learning environments, social negotiation, multiple perspectives, and self-awareness of the knowledge-production processes (Driscoll, 2000 ) . Any analysis of constructivism is dif fi cult because there is a great range of ideas and a variety of theoretical positions and differences in perception of the instructional implications of this basic tenet. In addition, “the
move away from the computational view brought about the move away from learning and cognition as the central focus of educational research in any form” (Winn, 2004 , p. 80). Principles deduced from constructive theories are numer-
ous and divergent. Though characteristics of constructive
learning as active, constructive, cumulative, collaborative,
situated and goal directed are canonical (Bednar et al., 1991 ;
De Corte, 2010 ; Shuell, 1988 ; Simons, 1993 ) , any learning
inherently shows this constructive character (Perkins, 1991 ) .
Given the divergence in interpretations of constructivism,
ranging from radical to moderate (Lowyck & Elen, 1993 ) , a
lack of precision in de fi ning principles for instructional inter-
ventions makes new prescriptions highly probabilistic (Winn,
1987 ) . Nevertheless, scholars derived constructive principles
to guide the design of so-called powerful learning environ-
ments. Driscoll ( 2000 ) , for example, formulates these prin-
ciples: (a) embed learning in complex, realistic and relevant
environments; (b) provide for social negotiation as an inte-
gral part of learning; (c) support multiple perspectives and
the use of multiple modes of representation; (d) encourage
ownership in learning; and (e) nurture self-awareness of the
knowledge construction process. Ertmer and Newby ( 1993 )
suggest these: (a) anchor learning in meaningful contexts;
(b) actively use what is learned; (c) revisit content at differ-
ent times, in rearranged contexts, for different purposes, and
from different conceptual perspectives; (d) develop pattern-
recognition skills presenting alternative ways of presenting
problems; and (e) present new problems and situations that
differ from the conditions of the initial instruction. Merrill
( 2002 ) elaborated fi rst principles that focus on knowledge
building and suggest that learning is promoted when: (a)
learners are engaged in solving real-world problems; (b)
existing knowledge is activated as a foundation for new
knowledge; (c) new knowledge is demonstrated to the
learner; (d) new knowledge is applied by the learner; and (e)
new knowledge is integrated into the learner’s world. These
three examples illustrate that generalized principles re fl ect
divergent fi ndings which renders operational advisement
almost impossible.
In contrast, the Jasper series (Cognition and Technology
Group at Vanderbilt, 1993 ) use concrete operationalization
of principles that involve video-based formats, narratives
with realistic problems, generative formats, embedded data
designs, problem complexity, pairs of related adventures,
and links across the curriculum. These seem to be descriptions
of speci fi c types of interactive instructional material
rather than theoretically derived and empirically validated
prescriptive principles (Elen, 1995 ) . The dif fi culty of detect-
ing and formulating principles for building constructive
learning reveals shortcomings in both theoretical precision
and convergent modeling. Jonassen and Reeves ( 1996 ) sug-
gest eliminating design principles and leaving design in the
hands of learners who use technologies as cognitive tools for
analyzing the world, accessing information, interpreting and
organizing their personal knowledge, and representing what
they know to others (i.e., learning by design or design-based
learning). Technologies such as databases, spreadsheets,
programming languages, visualization tools, micro-worlds,
and many others can be used to support such learning. What
is at issue is not constructivism as a theory but the learner’s
ability to cope with design complexity.
Socio-constructivism adheres to the viewpoint that human
activity is in fl uenced by affordances, artifacts, and other peo-
ple (Hewitt & Scardamalia, 1998 ) . In the broad framework
of a sociocultural approach, human activities are seen as
socially mediated (Dillenbourg, Baker, Blaye, & O’Malley,
1996 ; Lowyck & Pöysä, 2001 ) . Socio-constructivism adds
theoretical complexity while integrating learning, epistemo-
logical, sociological, anthropological, and educational theories
(Koschmann, 2001 ). Winn ( 2002 ) offers the following
principles for implementing the fi ndings of socio-construc-
tivism: (a) technology may sometimes be a necessary condi-
tion for the creation of learning communities but is never a
suf fi cient condition; (b) simply creating an interactive learn-
ing environment is not suf fi cient to bring about learning; (c)
practitioners should create a social context for learning in
technology-based learning environments; (d) effective learn-
ing communities often include experts from outside educa-
tion; (e) students should be encouraged, when appropriate, to
create or modify the learning environment; and (f) partner-
ships among students, teachers, and researchers should be
encouraged. However, these “should” statements are a source
of inspiration rather than an account of outcomes of research.
CSCL principles include these: (a) support educationally
effective peer interactions; (b) integrate different forms of
discourse; (c) focus students on communal problems of
understanding; (c) promote awareness of participants’ contributions; (e) encourage students to build on each other’s
work; and (f) emphasize the work on the community (Hewitt
& Scardamalia, 1998 ) . Again, these principles and sugges-
tions for application of theoretical fi ndings are framed in
general terms rather than in concrete links between theory,
fi ndings, principles, and prescriptions.
Evolutions in learning theory are translated into fi ndings and
principles that possibly guide the design of technological
tools. In most cases, it remains dif fi cult if not impossible to
detect a direct link between theory, and its operationalization
into technological tools or environments. The transitions
between theory, fi ndings, principles, and concrete imple-
mentations are problematic. Different research fi ndings lack
documentation of the transition steps between descriptive and prescriptive knowledge, which also caused problems in
building tools for automated instructional design (Spector,
Polson, & Muraida, 1993 ) . Most principles are formulated at
a general level, which supposes translation into very con-
crete situations, environments and tools. Consequently, the
expertise of designers, learners, and learner communities
will de fi ne effectiveness and ef fi ciency of these translation
The quest for understanding the links between learning and
things of learning started from the rather optimistic expecta-
tion that a close and natural relationship could be documented.
This expectation is suggested by the term
“educational technology.” However, in-depth scrutiny reveals
high complexity in both conceptualization and realization.
This led to the decision to represent the complexity in terms
of a limited set of observations to guide a critical appraisal of
the relationship between learning theories and technology.
These observations are subjective, based on selected sources,
and aim at further discussion. Within the limits of this
approach, a few main conclusions can be drawn.
Firstly, learning theories and technology show internal
and autonomous dynamics that lead toward mutual fertiliza-
tion. Their relationship is interdependent though not parallel,
and each can draw inspiration from the other. A tight empiri-
cal liaison, however, cannot be created. Ambitions of policy-
makers, researchers, and practitioners to innovate education
with new learning theories and powerful technologies,
yielded a myriad of isolated products, projects, and environ-
ments that were expected to impact education, learning and
learners in an effective and ef fi cient way. The aim to build
evidence-based knowledge about educational technology
mostly got stuck in idiosyncratic, divergent, and nebulous
frameworks. In contrast, interesting and worthwhile exam-
ples of links between learning theories and technology have
been found at a more fi ne-grained level of interaction in
which both learning principles and technological character-
istics are documented. These seldom led to valid theoretical
propositions that transcend the particularity of fi ndings or
Secondly, tuning learning theories to technology and vice
versa requires consistency and stability. Both domains show
intrinsic constraints that in fl uence modes of interaction. On
the one hand, learning theories can call for complex pro-
cesses that cannot be realized due to the limited capacities of
technology, as documented in the case of ITSs. On the other
hand, powerful technologies can be used for lower-level
learning goals, such as information delivery. In order to fos-
ter, the elaboration of a suitable conceptual framework that
focuses on interaction variables is urgently needed.
Thirdly, the relationship between learning theories and
technology is part of a complex educational system that calls
for synergy at the macro-, meso-, and micro-level. In addi-
tion, several parts of the system in fl uence the use of technol-
ogy for learning, which makes learning theories one of
several technology partners. Sociological, political, anthro-
pological, epistemological, fi nancial, economic, and organi-
zational and other issues play an important role in an
educational system. The question is if and to what degree an
interdisciplinary approach supports educational technology
theory and development. In the fi eld of educational technol-
ogy, isolation and balkanization of learning theories and
technologies hinder development of a linking discipline.
Fourthly, both learning theories and technology are empty
concepts when not connected to actors, such as instructional
designers, teachers, and learners. Many aspects of human
activity buffer the effectiveness and ef fi ciency of educational
technology. Deep understanding of learning theories and
technology as well as their relationship is a condition to activate
potential interplay and foster mutual fertilization.
Teachers and learners need metacognitive instructional
knowledge and motivation to tune their (mental) behaviors to
the nominal stimuli of the environment or to guide their own
process of learning in technology-enhanced learning envi-
ronments. To put it in a slogan, teachers and learners are co-
designers of their learning processes which affect
knowledge-construction and management as well as prod-
ucts that result from collaboration in distributed knowledge
Lastly the interplay between learning theories and tech-
nology needs a transition science. Learning theories deliver
descriptive fi ndings that fi ll the knowledge base of knowing
that , while educational technology, if not considered as tools
technology is a prescriptive fi eld that de fi nes knowing how ,
to use Ryle’s ( 1949 ) terminology. Instructional design as a ing how. Strange enough, learning theories and technology
become disconnected if instructional design does not consider
evolutions in learning theories. This is why strong
behaviorist principles that originated in early instructional
design hindered adaptation of models and principles to more
cognitive and constructivist approaches. Hopefully, evolu-
tions in learning theories and technologies will lead to more
coherence and synergy than has been illustrated with selec-
tions of the literature. This calls for a community that not
only designs and develops products and environments but
that invests in theory building through continuous re fi nement
of knowing that and knowing how to bring about synergy in
the complex and divergent fi eld of educational technology.
We shall not cease from exploration
And the end of all our exploring
Will be to arrive where we started
T.S. Eliot, Four quartets