In Restructuring Education Through Technology, I incorporated systems thinking to identify seven types of relationships in educational systems: teacher-student, student-content, student-context, teacher-content, teacher-context, content-context, and education system-environment relationships (Frick 1991). I now revisit these education system relations and discuss potential futures of education. The World Wide Web did not exist when I wrote the original treatise, nor did wireless smartphones and tablets, Google’s search engine, YouTube, Facebook, or Wikipedia. However, one important education system relationship should not change: the affective bonding between teachers and their students.
Avoid common mistakes on your manuscript.
As we enter the 2020 decade, 30 years have passed since I wrote a small book in 1990 called Restructuring Education Through Technology (Frick 1991). It was published about the same time as the World Wide Web first became available to the public, in August, 1991 (https://en.wikipedia.org/wiki/World_Wide_Web).
In the present article, I discuss important differences and similarities in education systems and technology in 1990 and 2020. I provide examples of education system affect-relations and system properties, then and now. I conclude with a strong philosophical stance on what is most important in the future.
Ludwig von Bertalanffy (1950a, 1950b, 1972) was a biologist who is well-known for his initial development of general system theory (GST). GST was intended as an organizational framework or paradigm applicable to many scientific fields, what von Bertalanffy originally considered a unifying science. Many others built upon von Bertalanffy’s foundational work, including Maccia and Maccia (1966), Wymore (1967), Mesarović (1972), Cornacchio (1972), Weinberg (1975), Berlinsky (1976), Lin (1999), and Ackoff and Emery (2017). Thompson (2006a, 2006b) briefly discussed the history of GST through the twentieth century, including the significant contribution of the SIGGS Theory Model (Maccia and Maccia, 1966). Predicate calculus was used to logically define approximately 80 terms. Thompson (2019) further refined these and other terms in Axiomatic Theories of Intentional Systems (ATIS).
Underlying these efforts is the fundamental goal of logical precision in use of terminology which describes general system, education, and education system. For example, system/negasystem flows were identified such as feedin, feedout, feedback, and feedthrough, and system structural properties were denoted such as size, complexity, strongness, flexibleness, wholeness, and hierarchical order. Further temporal properties included system openness, spillage, filtration, regulation, adaptiveness, compatibleness, and stress.
I present here a fundamental translation and description in natural language to support the main focus here. Think of this introduction to basic systems properties here as a guided tour of a building with many rooms. Readers can go as deep and detailed as they wish by following links to the web to explore further.
Figure 1 depicts an overview of fundamental systems concepts. The Appendix contains an abridged glossary of terms, which I have translated to natural language definitions.
The first distinction is between a system and its negasystem. A system is defined as “a set of components and a family of affect-relations” (Thompson 2019; https://aptac.sitehost.iu.edu/glossary/atisSystem.pdf). The negasystem is, literally, the not-system. In natural language, many people refer to system environment, but a distinction is made between toput and negasystem in SIGGS and ATIS. Toput is the system environment, which is part of the negasystem. Toput is what is available for selection by a system. If a system is open, then some of toput may later become system input. That temporal process is feedin. Some toput may not become input, what is termed system spillage. System filtration is a restriction of feedin.
Parallel terms include system fromput and system output. The process of stuff exiting the system is feedout. See Fig. 1 and the Appendix. System feedin and feedout are temporal processes. At time 1, some components are part of system toput, and at time 2 they are part of system input. System feedback is feedout followed at a later time by system feedin.
There are also structural affect-relations, which refer to the configuration of connectedness among component affect-relations, which are relatively stable over some period of time. Temporal affect-relations refer to the changing temporal configurations of event occurrences. In Fig. 1, system affect-relations are represented by the shape with arrows. In ATIS and SIGGS, mathematical digraphs (directed graphs) are used to characterize affect-relations. In order to keep matters relatively simple here, I will not digress into digraph theory, although it should be noted that it is fundamental to network analysis and complex systems, for example (e.g., see Jensen and Nielsen 2007; Brandes and Erlebach 2005). Noteworthy also is that components and their affect-relations also occur in the negasystem. Finally, due to space restrictions, more than 80 other important systems properties and their definitions are not discussed here. A broader description is available in Frick (2019) and on the Educology Website (2020) at: https://educology.iu.edu. Further details are provided in the ATIS glossary at https://aptac.sitehost.iu.edu/glossary/ and Thompson (2019).
An education system is but one kind of system. There are many kinds of systems, including systems of government, military, biology, economy, health care, etc., to name just a few. Some systems are intentional, while others are not (Thompson 2019). An education system is an intentional system. Intention refers to willing, i.e., trying to do, seeking a goal. The primary goal of an education system should be to guide student learning (Steiner 1981, 1988).
An education system is defined as an "intentional system that consists of at least one teacher and at least one student in a context” (Educology, 2020: https://educology.iu.edu/educationSystem.html).
Note the differences between Figs. 1 and 2. Now we are referring to an education system and an education negasystem. Note further the affect-relation shape in the education system, which now indicates relations among teachers, students, content, and contexts. Note also education system toput, which is the education system environment, which includes persons and things available for selection by an education system. Examples of affect-relations in education systems and negasystems are provided in Tables 1, 2, 3, 4, 5 and 6 below.
Finally, and most important, an education system is not confined to a place such as a school building with classrooms. This expansion is extremely important, and in fact is one of the main differences between 1990 and 2020 education systems. Some current education systems span the whole planet, and education system feedin and feedout occur through a network of networks referred to as the Internet and Worldwide Web.
An education system requires:
Note this conception of an education system consists of universals, not limited to time or place (https://educology.iu.edu/universals.html). Thus, education systems are not limited to schools or universities as we now know them.
These ideas are illustrated in Tables 1, 2, 3, 4, 5 and 6 , when comparing what was the same in 1990 and 2020, and what is new and different. The list is not exhaustive but is illustrative and parallels largely the organization in Frick (1991). Note that these are primarily examples of structural affect-relations in education systems and negasystems.
When I first drafted this article in late 2019, novel coronavirus was largely constrained to China. While I finalize the current article in June, 2020, a worldwide pandemic has been declared by the World Health Organization. In the U.S. many K-12 schools and colleges and universities have shut down their physical facilities (school buildings and campuses), and students and teachers were told to stay home during the past several months. Most teachers and professors have been suddenly thrown into a new context to do “emergency remote teaching” as a quick adaptation with little preparation (Hodges et al. 2020).
Nonetheless, while the physical facilities have been shut down, most education systems have not. The biggest structural change has been the education system context: students and teachers have remained physically located in their respective homes, and most student-teacher affect-relations have occurred via the Internet—at least for those who have access to the technologies required (e.g., through Zoom, e-mail, and course management systems, but see especially Table 6, last item in the right hand column). Those unfortunate have-nots have become disconnected from the education systems to which they belonged prior to the coronavirus pandemic (see ATIS property: https://aptac.sitehost.iu.edu/glossary/atisDisconnectedComponentsSet.pdf .)
The other obvious structural change since 1990 in student-context and teacher-context affect-relations is the current prevalence of digital devices and wireless connectivity to the Internet (wi-fi and cellular), as well as in content-context affect-relations (NCES 2018). Increasing amounts of content are now digitally encoded and stored in the Cloud (i.e., data centers accessible through the Internet: https://en.wikipedia.org/wiki/Cloud_computing).
One of the biggest challenges in 2020 is education system filtration (i.e., restriction of feedin). In 1990, publishers of textbooks, videos, and movies for education served as significant filters. What was available in education system environments (toput) was restricted through highly centralized control within the publishing industry and mass media outlets (see Thompson 2019, https://aptac.sitehost.iu.edu/glossary/atisCentralness.pdf).
In 2020, the size of content components (sheer numbers) in the education system environment (toput) is larger than that in 1990 by many orders of magnitude. There are so many more choices in 2020 of what is available in the education system environment when compared with 1990, especially on the Internet (e.g., see Bonk 2009; Mitra 2020; Waks 2016). For example, there were no websites in 1990, one in 1991, and now there are over 1.5 billion in June, 2020 (https://www.internetlivestats.com/total-number-of-websites/). YouTube reports that there are over 2 billion logged-in users each month, who watch over one billion hours of video daily (https://www.youtube.com/about/press/). The website developed at Indiana University on How to Recognize Plagiarism was viewed over 66 million times between 2016 and 2019, with nearly 600,000 students from 222 countries who had passed a certification test (https://plagiarism.iu.edu/recentChanges.html). The role of teachers in “selecting the best of culture” is still relevant, and now even more so with so much content available to students worldwide (Frick 1991, 2019).
Control of potential content for education systems has become much more decentralized in general, except in countries such as China where their government has barred access to “inappropriate” content by their citizens and students and teachers in Chinese education systems. The Chinese government has implemented a significant filter, restricting education system feedin (Xu et al. 2011).
On the other hand, in the U.S. and other countries which support more system openness, lack of effective filters has become a significant concern for educators, especially in K-12 schools (Bengfort 2019). Most of the vast resources on the Internet are unfiltered and uncurated—so there is a quality issue—it can often be difficult to determine what is trustworthy, safe, or accurate. For example, there are “fake news” publications, pornographic websites, and many other unsuitable materials (Kakutani 2018). Many K-12 school systems have attempted to filter toput, i.e., restrict Internet access at school by their students and teachers, by limiting which websites or Internet domains they can access (CIPA 2000).
Furthermore, students in 2020 can interact with their peers, friends, and others through social media—including Facebook, Instagram, Snapchat, Twitter, etc. Unfortunately, sexual predators and bullies also have access to these apps. Parents of students may try to put filters on their home networks, in order to restrict access to websites they consider to be inappropriate. But clever students can circumvent such attempts, especially when outside their home networks (e.g., see Hoffman 2019).
Space prevents further elaboration of important systems concepts in understanding similarities and differences in education systems in 1990 and 2020. However, I hope that the usefulness of universal systems concepts is patent to TechTrends readers.
I now want to address future education systems. Artificial intelligence (AI) and robot teachers could be in future education systems. We already have some examples of AI, such as Alexa and Siri. Just verbally ask a question, and in seconds there is an answer or a list of likely resources.
As mentioned earlier in Table 2, the Google search engine is likely the best current example of AI. Google bots roam the Web constantly, indexing billions of web pages. The Google Page Rank methodology effectively captures human judgement of what is important and relevant by examining what web page developers include as content and hyperlinks to other websites (e.g., see https://en.wikipedia.org/wiki/PageRank). Proprietary algorithms subsequently developed at Google have gone far beyond the original page rank method created by Brin and Page (1998) when doctoral students at Stanford University.
This raises the question that I posed over 20 years ago: “Will these globally interconnected, multimedia computer-television-telephone-stereo-radio systems largely replace teachers? Will these multimedia tutoring systems be intelligent enough to do so?” (Frick 1997, p. 108). First, there is a distinction between adaptive instruction vs. intelligent tutoring systems. We can already do adaptive instruction with today’s technologies. On the other hand, sentient machines as possible teachers is a futuristic scenario—one that is not within our reach with current technologies.
Adaptive instruction means essentially that what a student encounters next is customized or tailored according to what she or he has done in the past (e.g., Johnson and Sloan 2020). Historically, this was referred to as programmed instruction—i.e., structured by an instructional flowchart with decision nodes for branching (Heinich et al. 2002). In education systems terms, this is computer-mediated learning software that is designed to guide student learning (i.e., is instructed learning), and which has temporal properties of adaptiveness, feedback and feedthrough in order to increase compatibleness with student learners, and which has structural properties of flexibleness and wholeness. This we can do with present technologies.
Though we might not realize it, this kind of adaptiveness and customization has been going on as we surf the Web through techniques of storing browser cookies and digital fingerprinting (for an overview, see https://www.nytimes.com/2019/07/03/technology/personaltech/fingerprinting-track-devices-what-to-do.html). If you ever wondered how those running shoes you were shopping for on Amazon.com show up as an advertisement in Rolling Stone Magazine, here is what is going on in the “black box” you are using: When you visit some websites, they plant “tracking devices” on your computer, typically stored as cookies by your web browser (e.g., Chrome, Firefox, Safari, or Edge) and in more recent years by using your digital fingerprint and matching it against those recorded previously by websites.
In addition to cookies, our devices leave digital fingerprints that can uniquely and accurately identify us individually (Eckersley 2010). These device fingerprint and cookie data are in turn used by businesses for customizing advertisements tailored to match our unique profiles (Chen 2019). These kinds of data can also be used for more clandestine purposes—e.g., as tools for influencing democratic election outcomes, that is, how we vote—by determining which political messages to which we are individually more receptive, what Pariser (2011) calls a “filter bubble” (e.g., see Baer 2016).
In adaptive instruction, as we humans interact with computer software (now called “apps”), these programs can store records of what we have been doing while using the app. These records can in turn be used in conditional logic to govern what is displayed next by the computing device. Highly sophisticated pattern recognition can provide information that drives the conditional logic. This kind of adaptation can be done in computer tutorials, games, simulations, drill and practice, tests, etc. as part of an education system. Over the past five decades, I myself have engineered software programs that are adaptive, and have taught graduate students to do likewise (e.g., Enfield et al. 2012; Frick 1989; Frick, et al., 2020; Welch and Frick, 1993).
This kind of adaptation has long been done by human teachers as they become uniquely acquainted with their students and provide instruction that is appropriate for those students’ specific interests and ability to learn. Computer software today can emulate this kind of adaptation up to a point, the primary limitation being the “sensory apparatus” of the computing device (Frick 1997). Think of this kind of adaptation as a “smart” e-learning textbook, which is different for each student, depending on what she or he does and has learned thus far.
Next, I discuss so-called intelligent tutoring systems which are currently beyond our reach. Science fiction writers have envisioned some futuristic scenarios.
For example, Arthur C. Clarke (1968) included HAL as a central character in 2001: A Space Odyssey. The astronauts on the outer space mission conversed in natural language with HAL, played chess games, and queried HAL about ship operations, who was central to running the spaceship. In The Moon is a Harsh Mistress, Robert Heinlein (1966) described Mycroft Holmes (Mike), which was a sentient supercomputer who conversed with humans in natural language, could program itself, and was smart enough to help rebels living on the moon to overthrow the Lunar Authority, an arm of government controlled from planet earth. Mike and HAL were stationary computers.
In The Robots of Dawn (Azimov 1983), Daneel and Giskard were robots who moved around. Central to the plot, they helped a human detective solve a murder case—the death of another robot. Daneel was embodied in humanoid form, nearly indistinguishable from actual humans in form and function, including sex. And while Daneel was incredibly intelligent, Giskard, who was not humanoid, turned out to be even more sentient. This and other science fiction novels by Isaac Azimov were famous for introducing the “three laws of robotics”—prime directives for robots to follow: not harming human beings, following their orders, and self-preservation (https://en.wikipedia.org/wiki/Laws_of_robotics). In 3001: The Final Odyssey Clarke (1997) provided a further futuristic sci-fi scenario. “Braincaps” and “brainboxes” were sophisticated devices that could rapidly download knowledge to a human brain, essentially providing an instant education. Braincaps also allowed a human to have imaginary experiences—a virtual reality that was just as real as dreams, while interacting with others with their own virtual avatars.
In non-fiction, Ray Kurzweil (1999, 2005) envisions a future where humans and robots meld, transcending biology . He discusses six epochs of evolution, starting from physics and chemistry and ending with the universe waking up. He predicts that intelligence will evolve far beyond its current embodiment in organic forms, not unlike the underlying premise of Clarke’s first odyssey novel, centered around the mystery of intelligent, powerful monoliths buried on Earth’s moon and orbiting Jupiter.
These are but a few images of humankind’s future. Could future education systems turn out in any of these ways? Will we someday have humanoid robots, braincaps, or literally a universal mind?
Over 20 years ago, I discussed some of the major issues that will need to be addressed (Frick 1997), and many of these issues remain unresolved in 2020. More recently, Anderson and Rainie (2018) surveyed 979 experts on artificial intelligence who expressed their concerns and suggested solutions for our future. Notably, Erik Brynjolfsson, one of the experts surveyed, emphasized that, “… the right question is not ‘What will happen?’ but ‘What will we choose to do?’ We need to work aggressively to make sure technology matches our values.”
I, too, take a similar position. Rather than trying to predict what might be, I next discuss what education should be, taking a philosophical position.
Education is vital to society and our future.
Worthwhile education for everyone is the goal of making intrinsically goodand instrumentally goodeducationaccessible to everyone everywhere.
Every human being has a right to worthwhile education.
Why? To:
The conclusion reached by Greenspan and Benderly (1997) identifies the essence of the matter for human education:
Computers may be able to perform certain cognitive operations, even more effectively, and certainly faster, than humans. But unless they acquire the ability to experience and react to emotion, silicon chips will be unable to exercise intelligent discrimination…. What separates human intelligence from that of computers, robots, androids, and any other cyber-creatures we can imagine, is the fact that we possess a nervous system capable of—indeed specifically designed for—generating and evaluation of affect…. Unless and until we solve the problem of creating living cellular reactivity and affects, as well as the capacity to abstract patterns of affects, in an artificial form, no machine will think in a truly human way. (pp. 126-127)
Human teachers are essential to education. It is vital that teachers and students form affective bonds. This is the most important relationship in education systems that must be nurtured, no matter how technology might evolve. This is not an empirical claim from science or praxiology. It is a philosophic claim about what should be (Steiner 1981, 1988, 2009).
The conclusion at the end of Restructuring Education Through Technology is still relevant now and in the future, as it was then:
… the technology cannot select the best of culture for sharing with students.… In short, the technology cannot evaluate the worth of the content that we embody in the medium. That is our essential role as teachers. We must select the best of culture and share it with the next generation. (Frick 1991, p. 32)
For the more precise predicate calculus definitions, see Thompson (2019, https://aptac.sitehost.iu.edu/glossary/). The symbol, ‘=df’ means ‘equals by definition’.
Note that an education system should not be assumed to be synonymous with a physical place, such as a particular school system—e.g., a district in a city or county consisting of school buildings containing classrooms. Components and their affect-relations can be in different physical locations—e.g., students could be at their different homes, teachers at their homes, and the content on YouTube (via the Internet); teachers and students could communicate through the Internet (e.g., by texting, Zoom, Facebook, etc.). For further information and examples, see the Educology (2020) website: https://educology.iu.edu
This study was not funded.
