We appreciate the opportunity to comment on Vigor: Neuroeconomics of Movement Control, by Shadmehr and Ahmed. The authors' ideas are profoundly important and relate directly to analyses that guide work in our laboratories. We offer thoughts on the authors' foundational assumption that behavioral intensity (vigor) is proportional to the perceived value of outcomes driving behavior (incentives). The assumption is reasonable considering classical motivational thought and scholarship in related literatures (Heckhausen, Reference Heckhausen1991; Heckhausen & Heckhausen, Reference Heckhausen and Heckhausen2010). However, it is called into question by an influential contemporary theory by Brehm (e.g., Brehm and Self, Reference Brehm, Self, Rozenweig and Porter1989), which provides a distinctive and well-investigated perspective regarding relevant motivational processes (Gendolla, Wright, & Richter, Reference Gendolla, Wright, Richter and Ryan2012; Richter, Gendolla, & Wright, Reference Richter, Gendolla, Wright and Elliot2016; Wright, Reference Wright, Gollwitzer and Bargh1996).
Brehm's theory concerns motivational intensity – effort deployed at a point in time. The theory assumes that effort deployment is governed by a principle of conservation and distinguishes circumstances in which difficulty is fixed from ones in which difficulty is unfixed or unknown. Difficulty is fixed if performers can secure a desired outcome by attaining an established performance standard. Difficulty is unfixed if performers can secure outcomes of different value by attaining different performance standards. Difficulty is unknown if performers have no understanding of what will be required to secure a desired outcome.
Examples of the different difficulty circumstances might involve workers challenged to produce widgets. Difficulty would be fixed if the workers were told they could earn an amount of money for producing a set number within a particular timespan. Production lower than the set number would yield no money and production above would yield no more money than the designated amount. Difficulty would be unfixed if the workers were told they could earn different amounts of money depending on how many widgets they produced within the timespan. Difficulty would be unknown if the workers were told they could earn an amount of money for producing an unspecified number of widgets within the timespan.
How should effort be deployed in these circumstances? Brehm's theory argues that in all cases effort should be determined proximally (immediately) by appraisals of instrumental task difficulty. However, there are steps to consider and devils in detail. Consider workers presented the fixed widget challenge. Their first decision should be whether to meet the established performance standard, that is, generate the set number of widgets. Workers should decide affirmatively if they believe (a) they can succeed, and (b) the financial incentive justifies the effort that must be deployed. They should decide negatively if these conditions are not met. If the workers decide affirmatively, they should do what they believe is required, but – following the conservation principle – no more than what they believe is required. In other words, they should deploy effort that is consistent with their perception of task difficulty. If the workers decide negatively, they should do nothing. For them, effort deployment would violate the conservation principle either because it would be futile (fully wasteful) or because the value of effort deployed would exceed the perceived value of the incentive (relatively wasteful).
The resulting pattern of effort is striking. If incentive value is low, effort should correspond to difficulty at low difficulty levels, but be consistently low at moderate and high difficulty levels. If incentive value is moderate, effort should correspond to difficulty at low and moderate difficulty levels but be low at high difficulty levels. If incentive value is high, effort should correspond to difficulty at low, moderate, and high difficulty levels so long as success at the high levels is understood to be possible.
Notably, effort in this fixed circumstance should not be proportional to incentive value. It should comport with difficulty regardless of value where success is perceived as possible and worthwhile and be low regardless of value where success is perceived as impossible or excessively difficult considering incentive value. The only time incentive value should relate to effort is where it justifies requirements of a possible task when high, but not when low. Thus, for example, if some workers were offered a strong financial incentive for meeting a difficult, but possible, widget performance standard, and others were offered a weak financial incentive for doing so, effort should be high for the former group and low for the latter group.
Consider now workers presented the unfixed widget challenge. Their first decision should not be whether to meet an established performance standard. Rather, it should be what – if any – standard to pursue. Workers should decide considering what standards they can meet and what they can secure by performing at different levels. Possible associations between performance standard met and value of the outcome secured are variable. However, the association should commonly be linear, such that performers can secure steadily more value the better they perform. An example would be where workers were told they could earn 1 USD for each widget produced. Production at the first level would yield 1 USD. Production at the second would yield 2 USD. Production at the third would yield 3 USD. And so forth. In such a case, workers could be expected to select the most difficult performance standard that is (a) possible to meet, and (b) justified by the value of the outcome with which it is associated. Once workers select that standard, they should deploy effort correspondent to difficulty.
An insight associated with the linear widget example above relates to the value of the incentive at the first level of performance. The greater the value, the higher should be the maximum effort justified at each available performance level, with increasingly higher performance standard adoptions tending to follow. More specifically, as value increases, performers should tend to adopt increasingly higher performance standards until they reach the most difficult possible standard that is justified. Beyond this last point, increases in incentive value at the first level of performance should not alter standard adoption – and therefore effort – because performers will have attained their performance peak. Thus, in this special unfixed difficulty circumstance, we can expect behavioral vigor to reflect the value of the outcome driving behavior until performers are fully extended in terms of effort and can deploy no more.
Preceding points can be illustrated by considering scenarios in which workers were told they could earn more than 1 USD for each widget produced. Workers might view a 20-widget performance standard as excessively difficult (unjustified) if success would yield only 20 USD (1 USD × 20) but justified if success would yield 40 USD (2 USD × 20). Similarly, they might view a 25-widget performance standard as excessively difficult (unjustified) if success would yield 25 USD (1 USD × 25) but justified if success would yield 75 USD (3 USD × 25). If workers believe they can produce no more than 30 widgets and view that performance standard as justified by 120 USD (4 USD × 30), but nothing less than that, then increases in payment beyond 4 USD per widget should have no impact on standard choice or effort. To be clear, there are other details to consider, such as the difficulty of individual widget production. But the gist is the same for present purposes.
Consider finally workers presented the unknown widget challenge, the chance to earn an amount of money for producing an unspecified number of widgets. Once again, the first decision should not be whether to meet an established performance standard, because no standard is specified. Rather, the decision should be what – if any – standard to pursue and workers should decide considering what standards they can meet and what they can secure. A crucial point of note is that the likelihood of meeting the unspecified standard should increase directly with the standard met. Workers who aim high should have a better chance of meeting the unspecified standard than workers who aim low. This means that performers can secure a steadily more valuable outcome the better they perform. Outcome value at each performance level should be a function of the likelihood of meeting the standard at that level and the amount of money offered for meeting the unspecified standard. The greater the money offered, the greater should be value at each performance level.
As discussed above, when there is a linear association between performance standard met and value of the outcome secured, performers should (a) select the most difficult performance standard that is possible to meet and justified by the value of the outcome with which it is associated, and then (b) deploy the effort correspondent with difficulty. Consequently, we could expect workers presented the unknown widget challenge to do this. We also could expect their effort to reflect the value of the financial incentive up to the point that they attain their performance peak.
It is beyond the scope of this commentary to review empirical evidence for Brehm's effort arguments. Suffice it to say that it is considerable, with the most compelling coming from over 100 studies that have examined implications using focused cardiovascular responses as indices of effort (e.g., Richter et al., Reference Richter, Gendolla, Wright and Elliot2016). Studies include experiments that have manipulated the difficulty of behavioral challenges, the value of approach and avoidance incentives, and the nature (e.g., fixedness) of behavioral challenges – generating remarkable consistency in support.
A single example study is an experiment that examined the interactional influence of difficulty and incentive value operationalizing value in monetary terms (Eubanks, Wright, & Williams, Reference Eubanks, Wright and Williams2002). It presented participants over a series of work periods less or more difficult versions of a recognition memory task. More specifically, it presented character strings followed by a probe asking if a particular character was in the preceding string. Character string lengths were 1, 4, 7, 10, and 13, creating the different difficulty conditions. Instructions indicated that participants would earn chances to win a $10 prize (incentive value low) or a $100 prize (incentive value high) by maintaining a 90% success rate. Analysis of cardiovascular data collected during the work periods showed a difficulty × value interaction for heart rate. Heart rate increased steadily with difficulty among those who could win $100 but increased with difficulty only to a point for those who could win $10. As seen in Figure 1, heart rate corresponded with incentive value only at the highest difficulty level, presumably where effort was justified only if the high value incentive was available.
Insofar as Brehm's arguments have merit, they suggest that Shadmer and Ahmed's foundational assumption regarding the relationship between behavioral vigor and subjective value is warranted in some, but not all, performance circumstances. They also suggest that proportionality between vigor and value might be generated differently than the authors suppose. Specifically, it might be generated through a deliberative goal-setting process rather than through intrinsic neural linkages (Locke & Latham, Reference Locke and Latham2019). Means of reconciliation between Shadmer and Ahmed's perspective and Brehm's perspective are uncertain but could be available and would be well worth exploring.
You have
Access
We appreciate the opportunity to comment on Vigor: Neuroeconomics of Movement Control, by Shadmehr and Ahmed. The authors' ideas are profoundly important and relate directly to analyses that guide work in our laboratories. We offer thoughts on the authors' foundational assumption that behavioral intensity (vigor) is proportional to the perceived value of outcomes driving behavior (incentives). The assumption is reasonable considering classical motivational thought and scholarship in related literatures (Heckhausen, Reference Heckhausen1991; Heckhausen & Heckhausen, Reference Heckhausen and Heckhausen2010). However, it is called into question by an influential contemporary theory by Brehm (e.g., Brehm and Self, Reference Brehm, Self, Rozenweig and Porter1989), which provides a distinctive and well-investigated perspective regarding relevant motivational processes (Gendolla, Wright, & Richter, Reference Gendolla, Wright, Richter and Ryan2012; Richter, Gendolla, & Wright, Reference Richter, Gendolla, Wright and Elliot2016; Wright, Reference Wright, Gollwitzer and Bargh1996).
Brehm's theory concerns motivational intensity – effort deployed at a point in time. The theory assumes that effort deployment is governed by a principle of conservation and distinguishes circumstances in which difficulty is fixed from ones in which difficulty is unfixed or unknown. Difficulty is fixed if performers can secure a desired outcome by attaining an established performance standard. Difficulty is unfixed if performers can secure outcomes of different value by attaining different performance standards. Difficulty is unknown if performers have no understanding of what will be required to secure a desired outcome.
Examples of the different difficulty circumstances might involve workers challenged to produce widgets. Difficulty would be fixed if the workers were told they could earn an amount of money for producing a set number within a particular timespan. Production lower than the set number would yield no money and production above would yield no more money than the designated amount. Difficulty would be unfixed if the workers were told they could earn different amounts of money depending on how many widgets they produced within the timespan. Difficulty would be unknown if the workers were told they could earn an amount of money for producing an unspecified number of widgets within the timespan.
How should effort be deployed in these circumstances? Brehm's theory argues that in all cases effort should be determined proximally (immediately) by appraisals of instrumental task difficulty. However, there are steps to consider and devils in detail. Consider workers presented the fixed widget challenge. Their first decision should be whether to meet the established performance standard, that is, generate the set number of widgets. Workers should decide affirmatively if they believe (a) they can succeed, and (b) the financial incentive justifies the effort that must be deployed. They should decide negatively if these conditions are not met. If the workers decide affirmatively, they should do what they believe is required, but – following the conservation principle – no more than what they believe is required. In other words, they should deploy effort that is consistent with their perception of task difficulty. If the workers decide negatively, they should do nothing. For them, effort deployment would violate the conservation principle either because it would be futile (fully wasteful) or because the value of effort deployed would exceed the perceived value of the incentive (relatively wasteful).
The resulting pattern of effort is striking. If incentive value is low, effort should correspond to difficulty at low difficulty levels, but be consistently low at moderate and high difficulty levels. If incentive value is moderate, effort should correspond to difficulty at low and moderate difficulty levels but be low at high difficulty levels. If incentive value is high, effort should correspond to difficulty at low, moderate, and high difficulty levels so long as success at the high levels is understood to be possible.
Notably, effort in this fixed circumstance should not be proportional to incentive value. It should comport with difficulty regardless of value where success is perceived as possible and worthwhile and be low regardless of value where success is perceived as impossible or excessively difficult considering incentive value. The only time incentive value should relate to effort is where it justifies requirements of a possible task when high, but not when low. Thus, for example, if some workers were offered a strong financial incentive for meeting a difficult, but possible, widget performance standard, and others were offered a weak financial incentive for doing so, effort should be high for the former group and low for the latter group.
Consider now workers presented the unfixed widget challenge. Their first decision should not be whether to meet an established performance standard. Rather, it should be what – if any – standard to pursue. Workers should decide considering what standards they can meet and what they can secure by performing at different levels. Possible associations between performance standard met and value of the outcome secured are variable. However, the association should commonly be linear, such that performers can secure steadily more value the better they perform. An example would be where workers were told they could earn 1 USD for each widget produced. Production at the first level would yield 1 USD. Production at the second would yield 2 USD. Production at the third would yield 3 USD. And so forth. In such a case, workers could be expected to select the most difficult performance standard that is (a) possible to meet, and (b) justified by the value of the outcome with which it is associated. Once workers select that standard, they should deploy effort correspondent to difficulty.
An insight associated with the linear widget example above relates to the value of the incentive at the first level of performance. The greater the value, the higher should be the maximum effort justified at each available performance level, with increasingly higher performance standard adoptions tending to follow. More specifically, as value increases, performers should tend to adopt increasingly higher performance standards until they reach the most difficult possible standard that is justified. Beyond this last point, increases in incentive value at the first level of performance should not alter standard adoption – and therefore effort – because performers will have attained their performance peak. Thus, in this special unfixed difficulty circumstance, we can expect behavioral vigor to reflect the value of the outcome driving behavior until performers are fully extended in terms of effort and can deploy no more.
Preceding points can be illustrated by considering scenarios in which workers were told they could earn more than 1 USD for each widget produced. Workers might view a 20-widget performance standard as excessively difficult (unjustified) if success would yield only 20 USD (1 USD × 20) but justified if success would yield 40 USD (2 USD × 20). Similarly, they might view a 25-widget performance standard as excessively difficult (unjustified) if success would yield 25 USD (1 USD × 25) but justified if success would yield 75 USD (3 USD × 25). If workers believe they can produce no more than 30 widgets and view that performance standard as justified by 120 USD (4 USD × 30), but nothing less than that, then increases in payment beyond 4 USD per widget should have no impact on standard choice or effort. To be clear, there are other details to consider, such as the difficulty of individual widget production. But the gist is the same for present purposes.
Consider finally workers presented the unknown widget challenge, the chance to earn an amount of money for producing an unspecified number of widgets. Once again, the first decision should not be whether to meet an established performance standard, because no standard is specified. Rather, the decision should be what – if any – standard to pursue and workers should decide considering what standards they can meet and what they can secure. A crucial point of note is that the likelihood of meeting the unspecified standard should increase directly with the standard met. Workers who aim high should have a better chance of meeting the unspecified standard than workers who aim low. This means that performers can secure a steadily more valuable outcome the better they perform. Outcome value at each performance level should be a function of the likelihood of meeting the standard at that level and the amount of money offered for meeting the unspecified standard. The greater the money offered, the greater should be value at each performance level.
As discussed above, when there is a linear association between performance standard met and value of the outcome secured, performers should (a) select the most difficult performance standard that is possible to meet and justified by the value of the outcome with which it is associated, and then (b) deploy the effort correspondent with difficulty. Consequently, we could expect workers presented the unknown widget challenge to do this. We also could expect their effort to reflect the value of the financial incentive up to the point that they attain their performance peak.
It is beyond the scope of this commentary to review empirical evidence for Brehm's effort arguments. Suffice it to say that it is considerable, with the most compelling coming from over 100 studies that have examined implications using focused cardiovascular responses as indices of effort (e.g., Richter et al., Reference Richter, Gendolla, Wright and Elliot2016). Studies include experiments that have manipulated the difficulty of behavioral challenges, the value of approach and avoidance incentives, and the nature (e.g., fixedness) of behavioral challenges – generating remarkable consistency in support.
A single example study is an experiment that examined the interactional influence of difficulty and incentive value operationalizing value in monetary terms (Eubanks, Wright, & Williams, Reference Eubanks, Wright and Williams2002). It presented participants over a series of work periods less or more difficult versions of a recognition memory task. More specifically, it presented character strings followed by a probe asking if a particular character was in the preceding string. Character string lengths were 1, 4, 7, 10, and 13, creating the different difficulty conditions. Instructions indicated that participants would earn chances to win a $10 prize (incentive value low) or a $100 prize (incentive value high) by maintaining a 90% success rate. Analysis of cardiovascular data collected during the work periods showed a difficulty × value interaction for heart rate. Heart rate increased steadily with difficulty among those who could win $100 but increased with difficulty only to a point for those who could win $10. As seen in Figure 1, heart rate corresponded with incentive value only at the highest difficulty level, presumably where effort was justified only if the high value incentive was available.
Heart rate responses as a function of incentive value across five levels of memory task difficulty (from Eubanks et al., Reference Eubanks, Wright and Williams2002).
Insofar as Brehm's arguments have merit, they suggest that Shadmer and Ahmed's foundational assumption regarding the relationship between behavioral vigor and subjective value is warranted in some, but not all, performance circumstances. They also suggest that proportionality between vigor and value might be generated differently than the authors suppose. Specifically, it might be generated through a deliberative goal-setting process rather than through intrinsic neural linkages (Locke & Latham, Reference Locke and Latham2019). Means of reconciliation between Shadmer and Ahmed's perspective and Brehm's perspective are uncertain but could be available and would be well worth exploring.
Financial support
This study received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflict of interest
None.