Individual level cognitive theories of adult learning

According to Lehtinen and colleagues (2014), learning can be conceptualised as a complex system where learning processes on different levels interact but are not reducible to each other. One of the difficulties in formulating an exact definition of learning stems from the fact that there is a continuous process of adaptation and learning taking place without any learning intention or awareness. Thus, learning is not a single, well-defined event - it comprises a series of processes. On an individual level, learning refers to a constant developmental process of one’s thoughts and actions.

Simple learning processes have already been described in classical theories of conditioning. However, recent studies on brain processes and cell level chemical changes have re-described the elementary mechanisms of this type of learning (Gor-mezano, Prokasy, & Thompson, 2014). McClelland and his colleagues (McClelland, McNaughton, & O’Reilly, 1995, pp. 427-428) use the term “interleaved learning” to describe learning in which a “particular item is not learned all at once but is acquired only gradually, through a series of presentations interleaved with exposure to other examples from the domain”. This unconscious gradual learning is strongly shaping human behaviour and ways to interpret objects and events in the environment. It is a continuous, lifelong process, yet particularly crucial for young children’s learning of basic skills (e.g., intonation of mother tongue). This type of gradual formation of neural networks may be one of the key processes leading to tacit knowledge (see Chapter 9) typical of experienced adults. Studies also show that these gradual learning mechanisms based on high frequencies of repeated experiences can result in useful tacit knowledge but often they also lead to unfavourable learning results (Braithwaite, Pyke, & Siegler, 2017). This very basic form of learning is partly an independent process and partly interacts with the other forms of learning as described in the complex system approach.

There has been an important change in recent models explaining early cognitive development and learning. In contrast to the rationalist philosophical ideas developed by the major theories of learning of the 20th century, which claimed that human learning is based on a few general learning mechanisms, the recent models state that all the knowledge and skills an individual has are learned in interaction with the physical, social, and cultural environment. However, there is increasing evidence showing that evolution has equipped human beings with a rich variety of domain-specific innate predispositions called core cognition (Kanniloff-Smith, 2018; Kinzler & Spelke, 2007). These predispositions facilitate the acquisition of understanding in fields such as basic mathematics, geometry, and the physical and biological environment. According to the current viewpoint, innate domain specific dispositions do not directly lead to practical knowledge. For example, natural numbers are not actually “natural” but require a complex construction process based on innate predispositions, culturally mediated knowledge, and practice (Carey, 2009).

It is, however, an open question how domain-specific core cognition affects later learning in adulthood. Recent research on conceptual change (see Chapter 8) has shown that some of the early learning patterns influenced by the innate domain-specific predispositions are in conflict with the more advanced knowledge needed in scientific and complex technology environments. Some very recent evidence shows that early knowledge forms, particularly when related to the core cognition, do not disappear during later learning but co-exist with the more advanced scientific concepts - even in highly educated adults (Brault Foisy, Potvin, Riopel, & Masson, 2015; Obersteiner, Van Dooren, Van Hoof, & Verschaffel, 2013).

Current scientific explanations of conscious learning of conceptual knowledge and complex skills are based on models that describe the basic functioning of human memory. Already in the 1950s, Miller presented the seminal work on the limitations of human information processing in his article “The magical number seven, plus or minus two” (Miller, 1956). These early works marked the beginning of intensive research of the two different memory functions, working (or short-term) memory and long-term memory, and the relations between them in learning and other cognitive processes.

A comprehensive model of working memory has been presented by Baddeley (2007). According to that model, working memory consists of several interacting components such as a central executive, a visuospatial sketchpad, an episodic buffer, and a phonological loop. The central executive is an attention control system and three other components are short-term memory functions specialised for visual, episodic, and language information. The limited capacity of working memory in terms of time and dealing with simultaneous units means that the mechanisms of learning complex knowledge and skills require strategies which help learners adapt and overcome these limitations. The most common of these strategies is chunking or schema building, which means that the well-organised schemas and knowledge structures in long-term memory (prior knowledge, domain expertise) allow the activation of more complex units in working memory and lead to more effective learning and cognitive processing in new situations. These ideas were already developed in early works of cognitive science (e.g., Chase & Simon, 1973) and have later been elaborated further in many research traditions, including text comprehension (Kintsch & Mangalath, 2011), expertise research (Gong, Ericsson, & Moxley, 2015), and cognitive load research (Sweller, van Merrienboer, & Paas, 1998).

The cognitive load research tradition has carefully studied the optimal use of working memory in learning and the features of learning environments that support or hinder learning (Sweller et al., 1998). All learning tasks in intentional conceptual learning mean that the learner has to deal with units and relations between them in working memory. Increasing the elements means increasing the task-intrinsic cognitive load. Extraneous cognitive load refers to the unnecessary load resulting from an inappropriate design or presentation of the learning task. Germane cognitive load refers to the need of cognitive capacity that is related to creation and automatisation of schemas needed in dealing with the complex task. The conclusion for learning strategies and instructional design is to reduce extraneous cognitive load and focus the working memory capacity on intrinsic and germane cognitive load.

Ericsson and Kintsch (1995) have shown that very experienced adults can process tasks of their field of expertise in a way that overcomes the distinction between working and long-term memory. However, these limitations in memory also raise an interesting question of the average adult’s ability to handle many thinking systems at once (i.e., multiperspective and integrative thinking), as it is argued in current adult cognitive developmental models (see Part I of this book and Ghapter 2). This would mean that adults can compensate the gradually decreasing basic working memory capacity with increasing content knowledge. However, there are other studies which show that working memory capacity and content knowledge both contribute to the learning of new material but the content knowledge does not compensate the age-related decrease of working memory capacity (Hambrick & Engle, 2002).

Recent behavioural and neuroscience studies have focused particularly on the role of executive functions related to working memory, which are considered to be key processes in cognitive activity. Researchers have distinguished different components of executive functions. Of these, the widest interest has focused on updating, shifting, and inhibition. Updating refers to monitoring and the fast addition and deletion of working memory contents; shifting means flexibly changing between tasks; and inhibition means that the cognitive system overrides dominant responses so as to avoid distraction (Miyake & Friedman, 2012). Instead of looking separately at these different components, many researchers have focused on the common core of the functions, executive attention, and defined it as the ability to temporarily maintain goal-relevant information in memory (Schwaighofer, Fischer, & Buhner, 2015).

Executive functions are the basic mechanism for metacognition, selfregulations, and cognitive flexibility, and are thus crucial for the quality and effectiveness of learning. For example, in the research on conceptual change related to scientific concepts, inhibition has a crucial role (Brault Foisy et al., 2015). Executive functions are often related to intelligence, particularly fluid intelligence (Miyake & Friedman, 2012). Some researchers have seen the well-developed executive functions as one of the necessary requirements of more general wisdom, which is also partly linked to the concept of adult development and thus also to learning (see, Carlson et al., 2008; Chapter 2). Research has shown substantial individual differences and stability of the executive functions on the one hand, and malleability of them on the other hand (Miyake & Friedman, 2012; Schwaighofer et al., 2015).

Reviews about the effects of training for working memory and executive functions show that such training can enhance these basic functions but the effects are typically domain-specific and do not easily transfer to other cognitive domains (Schwaighofer et al., 2015). This is in line with the findings of expertise research which show that extensive training of specific skills can dramatically enhance specific cognitive control and attention regulation mechanisms which, according to the brain imaging findings, are related to a reduced activation of general executive functions (Hill & Schneider, 2006). There is a similar relationship between improving expertise and general abilities. Expertise studies show that after extensive deliberate practice leading to domain-specific expertise, the role of intelligence differences gradually fades as a predictor of achievement in that domain (Ericsson, 2014). Deliberate practice is the term proposed by Ericsson and his colleagues (Ericsson, Krampe, & Tesch-Romer, 1993) to describe systematic and intensive training which aims at purposefully developing different aspects of specific expertise. Research findings on deliberate practice are not only important for understanding a few exceptional experts but they also give convincing evidence about the plasticity of human neural and cognitive systems in adulthood (Hill & Schneider, 2006; Merrett, Peretz, & Wilson, 2013).

Mental representations and well-organised knowledge structures, which are developed through extensive deliberate practice, explain the superior achievement of experts (Ericsson & Pool, 2016). Concepts and methods to describe the development of mental representations, schemas, mental models, and knowledge structures are a major contribution of the cognitive research of the last 50 years (Johnson-Laird, 2010; Kintsch, 1998; Kintsch & Mangalath, 2011). Prior knowledge in terms of organised mental models and structures fundamentally affects conscious learning and construction of new knowledge, and the learning of conceptual knowledge can be described as changes in these models and structures. However, besides the conscious construction, human cognition develops also through the unconscious gradual formation of neural networks based on frequencies of features and events, and this can lead to learning which either supports the conscious conceptual construction or causes incorrect behaviour or beliefs that are conflicting with the conscious learning aims (see Braithwaite et al., 2017).