Cognitive Load Theory and Readability
Revised from a post published a year ago
Readable is a simple word with a straightforward meaning. An adjective, it is a good example of a self-referential word, that is, a word that reflects its meaning. Readable, the word, is, and is readily apparent to be, readable, demonstrating and illuminating its sense. Readable. Able to be read. Self-referential.
Pronunciation is another self-referential word, though of a different sort, one that puts the mouth through its paces in its performance, taking the mouth through a front round vowel (o), a mid central schwa (u), a tense high vowel (i) with the tongue shoved against the roof of the mouth gliding into an open, tongue-dropped, semi-rounded “a,” and resolving with another mid-central schwa.
Short is self-referential, exemplifying its meaning through its orthography. Long, however, is non-self-referential; shorter than short, long is not long. Simplicity shares meaning elements with readable; for a polysyllabic word, its simplicity is the essence of simplicity, embodied in its delicate pronunciation and its sturdy coherence.
Coherence is interesting, and I argue partly self-referential. To read the word, co- must mingle with -here, come together with other words coming from other places to here, joining together here. The process of reading the word demonstrates its meaning.
Clarity, simplicity, coherence, and ease seem appropriate on the list of descriptions of a readable text. You probably already know about studies showing how these qualities—clarity, simplicity, coherence, and ease—aren’t guaranteed by short sentences and short words. Forever, reading researchers have tried to count words and sentences and somehow be able to say something about whether a text is readable and by whom. Good luck with that.
Used in the context of specialized jargon, shape shifted from an adjective to a noun, readability changes its status from a self-referential word like readable to a heterological word A word is heterological if it does not possess the property that it denotes. “English” is a heterological word. “Polysyllabic” is polysyllabic, so it is not heterological.
Readability has nothing inherently to do with readable. Readability is a mathematical evaluation of linguistic surface features producing a scaled score signifying the grade level of achievement a reader must demonstrate to be expected to comprehend the text in question. Readability is a narrative, a process, not a quality. Strictly speaking, how readable a text is has little to do with its readability since readability is a narrative that reaches a statistical climax, not a quality.
Readability is interesting as a topic, in my opinion, but from my perspective of a teacher, it isn’t of much use. The primacy of lucidity, parsimony, and coherence in determining textual accessibility cannot be overstated, none of which is assessed via readability. From my teacher point of view, it would be fatally reductive to conflate lucidity, parsimony, and coherence with syntactic brevity or lexical familiarity as readability does.
That said, there is a meaningful difference between short and easy texts and long and hard texts. This difference is a direct consequence of readers’ individual experiences or lack thereof with similar texts and with the text’s content. In the real world where readable is what is actually readable, not a prediction, the only sure way to know is to let the reader try reading the text.
*****
Cognitive Load (CL) Theory developed by linguist John Sweller in the 1980s is concerned with demands made on cognitive activity during a learning activity. Lately, it has become a significant concept in discussions about AI and teaching and learning. Sweller’s model, which Sweller himself tinkered with over the decades into the 21st century, is highly relevant to understanding readability as a non-pedagogical concept.
The CL model was intended as a heuristic to explain how teaching activities can be designed to sustain learner focus and persistence by keeping the work—in this case, the difficulty of a text— at a manageable cognitive load. Too demanding results in breakdown; too simple results in superficial learning. If the goal is deep learning, i.e., building and integrating new schemas into existing knowledge networks, mental effort is required.
Mental effort explains why some texts seem harder to process than others, based on the cognitive resources required to understand them. Mental effort depends on a wide spectrum of factors, not the least of which are prior knowledge, and interest in the subject matter. Readability, then, is not assessable by way of the text, but by way of the reader’s experiences and motivation.
How information is presented in a text, how topics are structured and staged, is critical for readability, particularly for schema building. Preteaching organizational patterns in texts to reduce unnecessary intrinsic cognitive load directly impacts the reader’s experience processing the text.
*****
Sweller’s model defines three types of cognitive load which instructional strategies and designs can reduce or manage. The intrinsic cognitive load, the difficulty of the task, depends entirely upon each reader’s background and experience. In this regard, it makes more sense to assess the reader rather than the text if the intent is to predict cognitive load.
Assessing the reader can lead to preparation for the task before the task is engaged to reduce cognitive load. Preteaching core content vocabulary and concepts, providing advanced organizers, and walking learners through a reading assignment with an open-ended metacognitive discussion can reduce intrinsic load by providing tailor-made psychological tools.
Extraneous cognitive loads, those irrelevant mental demands that reduce the intensity of load available for reading the text, can be inadvertently intensified through instructional actions. Micromanaging the reading by requiring formula-driven behaviors (e.g., write down the five most important vocabulary words), for example, can divert reader focus from knowledge building to fulfilling the asks of the teacher. Reminding students or making salient a grade for the work can distort self-interested learning. Distractions of any sort even from the instructor can increase extraneous cognitive load.
*****
Collaborative resources involving peers during reading events can decrease anxiety and frustration as well as extraneous cognitive load by reducing the well-known fear of losing face that competition introduces. Extraneous load created by a competitive, individualist ethos and no opportunity to consult outside texts can be changed and may be a fruitful spot to explore AI bots. Collaborative resources can reduce extraneous cognitive load.
Germane load is the hard work of reading leading to increases in expertise, knowledge, and increased capacity to learn. Individual prior knowledge and experience are critical factors for this mental work, and students need to have self-assessed their own strengths and weaknesses in regard to the subject matter. Whether a text is readable is one thing. Whether the reader is prepared for the work that comes after the text is read is another.
Cognitive load addressing the integration of new knowledge and understanding with preexisting schemas, remodeling them as needed, creating new schemas, and making and naming links across schemas can be significantly strengthened through post-reading activities involving organizing, analyzing, synthesizing, and evaluating. Learners with fulsome opportunities to work in a learning community to objectify and represent information in their own fashions support germane cognition.
Cognitive Load Theory provides a basis for using the concept of readable text by enhancing preparation to read, psychological tools to use during reading, and guided instruction after reading toward knowledge consolidation.
A readable pedagogy isn't just about the words and the sentences on the page.