We applaud Morsella and colleagues for their original, innovative approach and welcome the uncommon, but from an evolutionary perspective very convincing, move (given that evolution operates on actions, not on thoughts) to link consciousness to action control. Many aspects of the theory invite empirical testing, which is likely to stimulate the field. However, we feel that to fully exploit its hypothesis-generating potential, the theory needs to be more specific with respect to the exact functionality that consciousness is assumed to have and to the timescale at which consciousness is thought to operate. We see two possible scenarios, which raise different questions.
The first scenario considers consciousness to operate online, that is, on the same short timescale in which the decision-making process operates. This would mean that consciousness not only emerges from/through action-selection conflict, but also operates on that same conflict. How could that work? Conflict solution requires the integration of information (including the conflicting parties and the goal the agent aims at). This fits with the integration consensus but raises the question: In which exact sense do Morsella et al. go beyond the global workspace theory of Baars (Reference Baars1988)? It also raises the question of how consciousness could influence the decision-making process involving conflict; is it not too slow to resolve conflict within the 450 msec window that typical responses in conflict tasks take, given that consciousness has been estimated to take hundreds of milliseconds to emerge (Dehaene et al. Reference Dehaene, Changeux, Naccache, Sackur and Sergent2006)? It is this feature, together with the lack of any empirical evidence for an online role of consciousness, that led Hommel (Reference Hommel2013) to doubt any online functionality of consciousness. Morsella et al. admit that there are “good reasons” to doubt that consciousness is involved in online action control, but they do not further address this as a problem for their own approach. Moreover, even if consciousness could emerge and operate on time, and even if it would allow for the access to distributed information that speeds up conflict solution, it is not clear why this integration function requires having a conscious experience. Why could a philosophical zombie without any conscious experience not possess such an integrative function? One case where online conscious experience appears to have a strong effect is addiction: conscious craving can overwhelm long-term considerations and lead to actions not consciously intended and later regretted (Kavanagh et al. Reference Kavanagh, Andrade and May2005; Wiers et al. Reference Wiers, Field, Stacy and Sher2014). However, this effect is at a longer timescale than experiments on online action control (seconds to minutes), and its effect is negative in view of the long-term goals of the individual. Hence, a positive example of conscious action control is still wanting.
The second scenario considers consciousness as an off-line function that does not operate during the ongoing perception-action event but prepares the agent for later events of the same sort. Rather than resolving the current conflict, consciously representing, reflecting, and communicating about the conflict and/or the solution could contribute to prevent the agent from encountering the conflict again, or at least to prepare her to deal with such conflicts more efficiently in the future. It may be no coincidence that the ability to communicate about an event (i.e., conscious report) is the most widely used technique to assess conscious representations in humans. Rather than merely a methodological convenience, the ability to communicate about conscious states may actually be the essence of what consciousness has evolved to achieve (Baumeister & Bargh Reference Baumeister, Bargh, Sherman, Gawronski and Trope2014; Baumeister & Masicampo Reference Baumeister and Masicampo2010; Hommel, Reference Hommel and Egnerin press). We may thus represent action conflicts consciously because that allows us to reflect on and communicate the existence of the conflict, our ways to deal with them, and the success of doing so. This allows for social learning and strategy transmission, but also for socializing the conflict.
Several theorists have argued that this indirect social role of consciousness has permitted humans to expand the number of individuals they interact with exponentially (Baumeister & Masicampo Reference Baumeister and Masicampo2010; Levitin Reference Levitin2014). Returning to our addiction example, by telling others about your goal to quit smoking you can mobilize them to help you when dealing with your urge to smoke the next time, this interaction serving as a reminder about your actual goals – you in essence externalize and socialize your goals and executive control functions. In treatment, you may further learn to “surf the urge” (Bowen & Marlatt Reference Bowen and Marlatt2009) and experience that the conscious urge will also descend when not acted upon. In addition, alternative strategies for weak moments are premeditated: if you often experience strong urges when stressed, it is important to prepare actions other than smoking (e.g., running, meditation, etc.) for upcoming stressful occasions, again externalizing future action control in a desirable way. Given the larger timescale of this operation mode, neither the slowness of conscious representation nor the absence of evidence for online functions of consciousness would be counter-arguments, and it would be obvious why conscious representations need to be conscious. To paraphrase Shariff et al. (Reference Shariff, Schooler, Vohs, Baer, Kaufman and Baumeister2008), although we may subjectively feel that our conscious will operates like a motorboat and we are steering where we want to go, the true operation of consciousness may be more indirect, like a sailing boat, in which we can learn to influence the boat's course in indirect ways, adjusting for the wind and the currents, which will eventually get us to our intended destination.
We would like to invite Morsella and colleagues to become more specific with regard to the timescale of conscious operations and their concrete functionality. This would not only strengthen the approach's potential to stimulate empirical research, but also strongly increase its well-deserved visibility and impact.
You have
Access
We applaud Morsella and colleagues for their original, innovative approach and welcome the uncommon, but from an evolutionary perspective very convincing, move (given that evolution operates on actions, not on thoughts) to link consciousness to action control. Many aspects of the theory invite empirical testing, which is likely to stimulate the field. However, we feel that to fully exploit its hypothesis-generating potential, the theory needs to be more specific with respect to the exact functionality that consciousness is assumed to have and to the timescale at which consciousness is thought to operate. We see two possible scenarios, which raise different questions.
The first scenario considers consciousness to operate online, that is, on the same short timescale in which the decision-making process operates. This would mean that consciousness not only emerges from/through action-selection conflict, but also operates on that same conflict. How could that work? Conflict solution requires the integration of information (including the conflicting parties and the goal the agent aims at). This fits with the integration consensus but raises the question: In which exact sense do Morsella et al. go beyond the global workspace theory of Baars (Reference Baars1988)? It also raises the question of how consciousness could influence the decision-making process involving conflict; is it not too slow to resolve conflict within the 450 msec window that typical responses in conflict tasks take, given that consciousness has been estimated to take hundreds of milliseconds to emerge (Dehaene et al. Reference Dehaene, Changeux, Naccache, Sackur and Sergent2006)? It is this feature, together with the lack of any empirical evidence for an online role of consciousness, that led Hommel (Reference Hommel2013) to doubt any online functionality of consciousness. Morsella et al. admit that there are “good reasons” to doubt that consciousness is involved in online action control, but they do not further address this as a problem for their own approach. Moreover, even if consciousness could emerge and operate on time, and even if it would allow for the access to distributed information that speeds up conflict solution, it is not clear why this integration function requires having a conscious experience. Why could a philosophical zombie without any conscious experience not possess such an integrative function? One case where online conscious experience appears to have a strong effect is addiction: conscious craving can overwhelm long-term considerations and lead to actions not consciously intended and later regretted (Kavanagh et al. Reference Kavanagh, Andrade and May2005; Wiers et al. Reference Wiers, Field, Stacy and Sher2014). However, this effect is at a longer timescale than experiments on online action control (seconds to minutes), and its effect is negative in view of the long-term goals of the individual. Hence, a positive example of conscious action control is still wanting.
The second scenario considers consciousness as an off-line function that does not operate during the ongoing perception-action event but prepares the agent for later events of the same sort. Rather than resolving the current conflict, consciously representing, reflecting, and communicating about the conflict and/or the solution could contribute to prevent the agent from encountering the conflict again, or at least to prepare her to deal with such conflicts more efficiently in the future. It may be no coincidence that the ability to communicate about an event (i.e., conscious report) is the most widely used technique to assess conscious representations in humans. Rather than merely a methodological convenience, the ability to communicate about conscious states may actually be the essence of what consciousness has evolved to achieve (Baumeister & Bargh Reference Baumeister, Bargh, Sherman, Gawronski and Trope2014; Baumeister & Masicampo Reference Baumeister and Masicampo2010; Hommel, Reference Hommel and Egnerin press). We may thus represent action conflicts consciously because that allows us to reflect on and communicate the existence of the conflict, our ways to deal with them, and the success of doing so. This allows for social learning and strategy transmission, but also for socializing the conflict.
Several theorists have argued that this indirect social role of consciousness has permitted humans to expand the number of individuals they interact with exponentially (Baumeister & Masicampo Reference Baumeister and Masicampo2010; Levitin Reference Levitin2014). Returning to our addiction example, by telling others about your goal to quit smoking you can mobilize them to help you when dealing with your urge to smoke the next time, this interaction serving as a reminder about your actual goals – you in essence externalize and socialize your goals and executive control functions. In treatment, you may further learn to “surf the urge” (Bowen & Marlatt Reference Bowen and Marlatt2009) and experience that the conscious urge will also descend when not acted upon. In addition, alternative strategies for weak moments are premeditated: if you often experience strong urges when stressed, it is important to prepare actions other than smoking (e.g., running, meditation, etc.) for upcoming stressful occasions, again externalizing future action control in a desirable way. Given the larger timescale of this operation mode, neither the slowness of conscious representation nor the absence of evidence for online functions of consciousness would be counter-arguments, and it would be obvious why conscious representations need to be conscious. To paraphrase Shariff et al. (Reference Shariff, Schooler, Vohs, Baer, Kaufman and Baumeister2008), although we may subjectively feel that our conscious will operates like a motorboat and we are steering where we want to go, the true operation of consciousness may be more indirect, like a sailing boat, in which we can learn to influence the boat's course in indirect ways, adjusting for the wind and the currents, which will eventually get us to our intended destination.
We would like to invite Morsella and colleagues to become more specific with regard to the timescale of conscious operations and their concrete functionality. This would not only strengthen the approach's potential to stimulate empirical research, but also strongly increase its well-deserved visibility and impact.