Elite athletes perform visually guided actions that test the limits of what is humanly possible. Think, for instance, of a baseball batter such as Aaron Judge who has 400 ms to hit Shohei Ohtani’s fastball. Elite athletes such as Judge train tirelessly to overcome these temporal constraints, and therefore, fast ball-sports provide some of the best possible evidence for understanding the limits of human action-guidance (Loffing, Reference Loffing2017). What can be learned from sports that helps to better understand the idea of unconscious action-guidance?
Like many theorists, Michel and Fleming distinguish between conscious and unconscious seeing. They argue that we can experimentally determine a lower bound for the time course of conscious seeing. Given this lower bound, if some instances of seeing required for action are faster than this lower bound, then that seeing must be unconscious, occurring before consciousness has time to emerge. Presumably, the seeing needed to hit a fastball would be unconscious, and this fits with an influential picture of human cortical vision that divides the dorsal and ventral streams as unconscious and conscious, respectively (Milner & Goodale, Reference Milner and Goodale1995).
What is this test of consciousness that we can use to determine the lower bound? Effectively, it’s reportability, either verbally or by some proxy (alternatively, cognitive accessibility, but the questions we raise will remain). Unconscious vision is not reportable even if it guides behavior such as hitting. Notice that in either the unconscious or conscious case, vision is supporting intentional behavior. Why is one intentional action the litmus test for consciousness while the other is not (Wu, Reference Wu2020)?
Imagine a different test, namely, the presence of visually guided intentional behavior. Intentional action can be more broadly or narrowly interpreted, but it must at least exclude automatic reflexes while also including more than report (or behavior tied to cognitive access). While it is striking that humans sometimes cannot report or remember the stimuli to which they act with near normal responses, it is a further theoretical commitment, not obligatory, that reportability (cognitive access) should be the test for consciousness, such that it rules out other seemingly intentional visually guided actions. Michel and Fleming’s time course argument depends on prioritizing certain behaviors over others.
We note this because the issues raised by actions like batting look different if we drop the conscious versus unconscious framing. Instead, we can talk about seeing that serves quick motor responses and seeing that serves slower cognitive processes, or revamping Milner and Goodale’s distinction, seeing-for-movement and seeing-for-thought. This allows us to ask an important question whose interest does not require the paradoxical merging of zombie and conscious visual streams in normal behavior (Abernethy & Mann, Reference Abernethy and Mann2008; Wu, Reference Wu2013): how do fast seeing-for-movement and slow seeing-for-thought work together to explain elite performance?
Consider, for instance, the results of studies that have deliberately blurred the vision of athletes, particularly when hitting a cricket ball. Batters are typically terrified when the blur is introduced, having very little expectation that they will be able to see the ball, let alone hit it. Yet, results show that their hitting performance is surprisingly resilient to considerable amounts of blur (Mann et al., Reference Mann, Ho, De Souza, Watson and Taylor2007, Mann, Abernethy, & Farrow, Reference Mann, Abernethy and Farrow2010a, Reference Mann, Abernethy and Farrow2010b), with one study showing that batters needed to be blurred to the point of legal blindness (20/200 or “10% vision”) before their performance even started to decrease (Mann et al., Reference Mann, Ho, De Souza, Watson and Taylor2007). These findings have been interpreted based on a dissociation between the ventral and dorsal visual stream. The ventral stream required for seeing-for-thought relies on high-quality vision that is detailed and colorful, while the dorsal stream required for seeing-for-movement relies on more primitive visual information that is blurrier but sensitive to movement and contrast (Livingstone & Hubel, Reference Livingstone and Hubel1988; Milner & Goodale, Reference Milner and Goodale1995). It has been suggested that the blur has little impact on hitting – up to a point – because the dorsal stream relies on already impoverished visual information.
Yet, seeing-for-thought (i.e., the ventral stream) does not sit idle when performing visually guided actions. Elite athletes often attest to using information available in short time-scales for action guidance. Baseball batters frequently report using the rotation of the seams on the ball during flight to guide whether to swing, and if so, where to swing towards (Clutter et al., Reference Clutter, Do, Zimmerman and Fogt2024). Similarly, an internationally renowned cricket batter reported that it could not be true that his gaze was behind the ball at the moment he hit it – as was the dogma at the time – because he could clearly see the markings on the ball during flight, and, in particular, just before/at contact. Testing revealed that he did indeed direct his gaze towards the ball at the moment he hit it (Mann, Spratford, & Abernethy, Reference Mann, Spratford and Abernethy2013). An even more successful teammate said he thought it was impossible to watch the ball at the moment it is hit, but testing showed that his gaze, too, was on the ball when he hit it. But he did report trying to optimize his technique by “watching the ball emerge from under my bat.”
This leads to the further consideration that explicit strategies drawing on vision-for-thought can be used to guide behavior and improve hitting abilities. Hitters who are not fully following the ball with their eyes are sometimes coached to report during training which part of the bat hits the ball when they make contact (e.g., whether the ball hits the close or distant part of the bat). This explicit task has been found to result in tennis players tracking the ball longer and ultimately improving their hitting accuracy. Hitting arguably integrates seeing-for-movement and seeing-for-thought to support elite performance (Sève et al., Reference Sève, Saury, Theureau and Durand2002).
Michel and Fleming might find these points congenial, for they are free to translate our framing back into the conscious versus unconscious vision. Yet to do so, they must make a strong theoretical stance on what behavior indicates consciousness. We think the issues raised without taking this controversial stance are just as interesting, namely that seeing can serve different functions, and that in normal agents, seeing must be integrated for intentional action.
You have
Access
Elite athletes perform visually guided actions that test the limits of what is humanly possible. Think, for instance, of a baseball batter such as Aaron Judge who has 400 ms to hit Shohei Ohtani’s fastball. Elite athletes such as Judge train tirelessly to overcome these temporal constraints, and therefore, fast ball-sports provide some of the best possible evidence for understanding the limits of human action-guidance (Loffing, Reference Loffing2017). What can be learned from sports that helps to better understand the idea of unconscious action-guidance?
Like many theorists, Michel and Fleming distinguish between conscious and unconscious seeing. They argue that we can experimentally determine a lower bound for the time course of conscious seeing. Given this lower bound, if some instances of seeing required for action are faster than this lower bound, then that seeing must be unconscious, occurring before consciousness has time to emerge. Presumably, the seeing needed to hit a fastball would be unconscious, and this fits with an influential picture of human cortical vision that divides the dorsal and ventral streams as unconscious and conscious, respectively (Milner & Goodale, Reference Milner and Goodale1995).
What is this test of consciousness that we can use to determine the lower bound? Effectively, it’s reportability, either verbally or by some proxy (alternatively, cognitive accessibility, but the questions we raise will remain). Unconscious vision is not reportable even if it guides behavior such as hitting. Notice that in either the unconscious or conscious case, vision is supporting intentional behavior. Why is one intentional action the litmus test for consciousness while the other is not (Wu, Reference Wu2020)?
Imagine a different test, namely, the presence of visually guided intentional behavior. Intentional action can be more broadly or narrowly interpreted, but it must at least exclude automatic reflexes while also including more than report (or behavior tied to cognitive access). While it is striking that humans sometimes cannot report or remember the stimuli to which they act with near normal responses, it is a further theoretical commitment, not obligatory, that reportability (cognitive access) should be the test for consciousness, such that it rules out other seemingly intentional visually guided actions. Michel and Fleming’s time course argument depends on prioritizing certain behaviors over others.
We note this because the issues raised by actions like batting look different if we drop the conscious versus unconscious framing. Instead, we can talk about seeing that serves quick motor responses and seeing that serves slower cognitive processes, or revamping Milner and Goodale’s distinction, seeing-for-movement and seeing-for-thought. This allows us to ask an important question whose interest does not require the paradoxical merging of zombie and conscious visual streams in normal behavior (Abernethy & Mann, Reference Abernethy and Mann2008; Wu, Reference Wu2013): how do fast seeing-for-movement and slow seeing-for-thought work together to explain elite performance?
Consider, for instance, the results of studies that have deliberately blurred the vision of athletes, particularly when hitting a cricket ball. Batters are typically terrified when the blur is introduced, having very little expectation that they will be able to see the ball, let alone hit it. Yet, results show that their hitting performance is surprisingly resilient to considerable amounts of blur (Mann et al., Reference Mann, Ho, De Souza, Watson and Taylor2007, Mann, Abernethy, & Farrow, Reference Mann, Abernethy and Farrow2010a, Reference Mann, Abernethy and Farrow2010b), with one study showing that batters needed to be blurred to the point of legal blindness (20/200 or “10% vision”) before their performance even started to decrease (Mann et al., Reference Mann, Ho, De Souza, Watson and Taylor2007). These findings have been interpreted based on a dissociation between the ventral and dorsal visual stream. The ventral stream required for seeing-for-thought relies on high-quality vision that is detailed and colorful, while the dorsal stream required for seeing-for-movement relies on more primitive visual information that is blurrier but sensitive to movement and contrast (Livingstone & Hubel, Reference Livingstone and Hubel1988; Milner & Goodale, Reference Milner and Goodale1995). It has been suggested that the blur has little impact on hitting – up to a point – because the dorsal stream relies on already impoverished visual information.
Yet, seeing-for-thought (i.e., the ventral stream) does not sit idle when performing visually guided actions. Elite athletes often attest to using information available in short time-scales for action guidance. Baseball batters frequently report using the rotation of the seams on the ball during flight to guide whether to swing, and if so, where to swing towards (Clutter et al., Reference Clutter, Do, Zimmerman and Fogt2024). Similarly, an internationally renowned cricket batter reported that it could not be true that his gaze was behind the ball at the moment he hit it – as was the dogma at the time – because he could clearly see the markings on the ball during flight, and, in particular, just before/at contact. Testing revealed that he did indeed direct his gaze towards the ball at the moment he hit it (Mann, Spratford, & Abernethy, Reference Mann, Spratford and Abernethy2013). An even more successful teammate said he thought it was impossible to watch the ball at the moment it is hit, but testing showed that his gaze, too, was on the ball when he hit it. But he did report trying to optimize his technique by “watching the ball emerge from under my bat.”
This leads to the further consideration that explicit strategies drawing on vision-for-thought can be used to guide behavior and improve hitting abilities. Hitters who are not fully following the ball with their eyes are sometimes coached to report during training which part of the bat hits the ball when they make contact (e.g., whether the ball hits the close or distant part of the bat). This explicit task has been found to result in tennis players tracking the ball longer and ultimately improving their hitting accuracy. Hitting arguably integrates seeing-for-movement and seeing-for-thought to support elite performance (Sève et al., Reference Sève, Saury, Theureau and Durand2002).
Michel and Fleming might find these points congenial, for they are free to translate our framing back into the conscious versus unconscious vision. Yet to do so, they must make a strong theoretical stance on what behavior indicates consciousness. We think the issues raised without taking this controversial stance are just as interesting, namely that seeing can serve different functions, and that in normal agents, seeing must be integrated for intentional action.
Financial support
The authors have no funding to declare.
Competing interests
The authors have no competing interests to declare.