Some constructs in motivation science are certainly underdeveloped and some motivation researchers may work with underspecified constructs, as suggested by Murayama and Jach (M&J). However, this is not indicative of a general problem in motivation science. Many motivation theories focus on specific mechanisms underlying motivated behavior and thus have already adopted the computational process perspective that M&J call for.

Murayama and Jach criticize the use of high-level motivational constructs in psychology, urging psychologists to “unpack” the black box. These constructs are alleged to be “psychological constructions” with no causal powers of their own. I argue that this view is mistaken, and that high-level motivational constructs are causal even when unpacked in terms of underlying computational, algorithmic, and implementational processes.

Though we see the potential for benefits from the development of process-oriented approaches, we argue that it falls prey to many of the same critiques raised about the existing construct level of analysis. The construct-level approach will likely dominate motivation research until we develop computational models that are not only accurate, but also broadly usable.

We argue that the target article's computational/reductionistic approach to motivation is insufficient to explain the energization of human behavior, because such explanation requires broad consideration of “what people are trying to do.” We illustrate what is gained by retaining (rather than jettisoning) higher-order motivation constructs and show that the authors’ approach assumes, but fails to name, such constructs.

There are indeed questionable motivation constructs in psychology. The diagnosis and proposed remedies in the target article both neglect the crucial consideration that all tendencies to behaviour compete for the same finite set of degrees of freedom. Action selection also has irreducibly economic aspects which should constrain motivation constructs and already inform healthy research programmes.

Integrating the predictive processing framework into our understanding of motivation offers promising avenues for theoretical development, while shedding light on the computational processes underlying motivated behavior. Here we decompose expected free energy into intrinsic value (i.e., epistemic affordance) and extrinsic value (i.e., instrumental affordance) to provide insights into how individuals adapt to and interact with their environment.

Understanding the psychological computations underlying motivation can shed light onto motivational constructs as emergent phenomena. According to Murayama and Jach, reward-learning is a key candidate mechanism. However, there's no such thing as a free lunch: Not only benefits (like reward), but also costs inherent to motivated behaviors (like effort, or uncertainty) are an essential part of the picture.

Murayama and Jach critically evaluate the idea that motivation is a dynamic that determines behavior and propose alternatively that it might be an emergent property that people construe through perceived regularities in experience and action. The critique has value but fails to appreciate the progress that has been made in moving beyond the idea of which the authors are critical.

Although higher-level constructs often fail to explain the mechanisms underlying motivation, we argue that purely mechanistic approaches have limitations. Lower-level neural data help us identify “biologically plausible” mechanisms, while higher-level constructs are critical to formulate measurable behavioral outcomes when constructing computational models. Therefore, we propose that a multi-level, multi-measure approach is required to fully unpack the black box of motivated behavior.

The target article argued that motivation constructs are treated as black boxes and called for work that specifies the mental computational processes underlying motivated behavior. In response to critical commentaries, we clarify our philosophical standpoint, elaborate on the meaning of mental computational processes and why past work was not sufficient, and discuss the opportunities to expand the scope of the framework.