Psychologist Gordon Logan at Vanderbilt University will use the deceptively simple task of typing to study "cognitive control," that is, how we plan and execute complex actions. Skilled typists appear to effortlessly use an external script (when copying) or an internally generated plan (when composing) to choose the appropriate movement in the form of a key-press 5-6 times per second. However, in a recent experiment, Logan showed that despite such virtuosity, skilled typists do not even know which hand they are using to depress a key at any given moment. The demonstration is clever and simple: try typing a sentence using only your left hand and typing only those letters normally typed with the left hand. Not surprisingly, this instruction slows people down, but it also renders them greatly less accurate (5-6 times more errors are produced), indicating that skilled typists are simply not aware of what they are doing at the most basic level. Logan's theory of cognitive control as applied to typing assumes that proficient typing is the result of the coordination between two "nested feedback loops." The "outer" loop determines the sequence of words to be typed and the "inner" loop translates the words into the appropriate keystrokes. This explains the paradox that skilled typists are not aware of something as basic as which hand they are using: the inner loop automatically and effortlessly determines the sequence of key presses needed to produce the sequence of words planned by the outer loop, but the outer loop does not know or care how the inner loop performs its separate task. This project will examine three aspects of this double-loop model of cognitive control: awareness (how much does the outer loop know about what the inner loop is doing?), structure (do words or some other level of organization, such as letters or phrases, serve as the interface between the two loops?), and feedback (the outer loop is hypothesized to rely on visual feedback while the inner loop is hypothesized to rely on kinesthetic feedback).
A better understanding of the mechanisms of cognitive control could lead to improvements in the way complex skills are taught and practiced. In the domain of typing, a detailed understanding of the process could have implications for the design of keyboards and other text-input devices. With the proliferation of computers in modern culture, typewriting is an essential skill in the workplace, in schools and universities, and in leisure activities. This project will also have an important impact on the broader literature on cognitive control, which is embroiled in a controversy over hierarchical processing. The outer-loop inner-loop theory provides a new perspective on research on automatic and controlled cognitive processes, which are often regarded as opposites. This project shifts the research focus from whether processes are controlled to how processes are controlled.
The main goal of the project was to document and critically evaluate evidence for hierarchical control of cognitive skills like reading, speaking, playing music and sports, in order to resolve longstanding controversies about the nature of cognitive control in cognitive science and cognitive neuroscience. The project focused on skilled typewriting, distinguishing between higher-level processes that generate ideas and translate them into a series of words to be typed and lower-level processes that translate words into a series of keystrokes. The strategy guiding the research was to dissociate higher-level and lower-level processes in three respects, addressed in three specific aims. The first aim was to determine the conditions under which higher-level processes were and were not aware of the details of lower-level processes. This aim was addressed in experiments that required skilled typists to type texts but only type the letters assigned to one hand (e.g., the right hand). Typists found this extremely difficult, slowing their typing speed from 80 words per minute to 14 words per minute, suggesting that higher-level processes are not aware of which hand types which letter. The second aim was to test a key assumption of hierarchical processing, that a single higher-level unit is broken down into many lower-level units. This aim was addressed in behavioral and electrophysiological experiments that demonstrated that all of the keystrokes in a word were active simultaneously in the motor system as typists typed words. The third aim was to distinguish the kinds of feedback that higher- and lower-level processes rely on. This aim was accomplished in experiments that manipulated the kind of feedback displayed on the screen as skilled typists typed words, sometimes inserting errors in words typists typed correctly and sometimes correcting errors in words that typists typed erroneously. Typists reports of their errors reflected what was displayed on the screen. They took credit for corrected errors and took the blame for inserted errors in words they typed correctly. Their fingers knew the truth, slowing down after the errors they actually committed whether or not we corrected them, and not slowing down after errors we inserted. The intellectual merit of the project was to provide insights into the nature of hierarchical control in cognitive skills, generalizing the results from skilled typewriting to other cognitive skills like reading, speaking, playing music, and playing sports. These insights are important in basic and applied research on cognitive skills and in evaluating specific theories of typewriting. The insights are important more generally because they resolve long-standing controversies about hierarchical control in cognitive science and cognitive neuroscience. The results provide clear evidence for hierarchical control in skilled typewriting and provide methods for addressing the controversy in many other domains. The broader impact of the project is to provide basic science on the hierarchical nature of cognitive control, which is important in theory and in practice. The focus on skilled typewriting is important because typing that has become increasing prevalent in the modern computer age, permeating every aspect of life, from the workplace to recreational activities. Understanding the nature of the skill and the factors that promote it will impact many people’s daily lives. The basic science will support evidence-based educational practices and training procedures. It provides insights into people’s everyday behaviors that may pique their interest in science and inspire some to pursue careers in science, technology, education, and mathematics. The research projects engaged several undergraduate students, graduate students, and postdoctoral fellows, providing them with training on research methods, including designing, programming, executing, and analyzing experiments and writing reports of the research for publication.
