The commentary discusses three components of the target proposal: (1) analogy as a host of phenomena, (2) relations as transformations, and (3) analogy as priming. The commentary argues that the first component is potentially productive, but it has yet to be fully developed, whereas the second and third components do not have an obvious way of accounting for multiple counterexamples.

The target article by Leech et al. presents a compelling computational theory of analogy-making. However, there is a key difficulty that persists in theoretical treatments of analogy-making, computational and otherwise: namely, the lack of a detailed account of the neurophysiological mechanisms that give rise to analogy behavior. My commentary explores this issue.

If connectionist computational models explain the acquisition of complex cognitive skills, errors in such models would also help explain unusual brain activity such as in creativity – as well as in mental illness, including childhood onset problems with social behaviors in autism, the inability to maintain focus in attention deficit and hyperactivity disorder (ADHD), and the lack of motivation of depression disorders.

Leech et al. present a connectionist algorithm as a model of (the development) of analogizing, but they do not specify the algorithm's associated computational-level theory, nor its computational complexity. We argue that doing so may be essential for connectionist cognitive models to have full explanatory power and transparency, as well as for assessing their scalability to real-world input domains.

Abstract thought, in general, and – reasoning by analogy, in particular, have been said to reside at the very summit of human cognition. Leech et al. endeavor to comprehend the development of analogous thinking in human beings. Applying Leech et al.'s general approach to the evolution of analogical behavior in animals might also prove to be of considerable value.

In this response, we reiterate the importance of development (both ontogenetic and phylogenetic) in the understanding of a complex cognitive skill – analogical reasoning. Four key questions structure the response: Does relational priming exist, and is it sufficient for analogy? What do we mean by relations as transformations? Could all or any relations be represented as transformations? And what about the challenge of more complex analogies? In addressing these questions we bring together a number of supportive commentaries, strengthening our emergentist case for analogy (in particular with insights from comparative psychology), and review new supportive evidence. We also rebut those commentaries that ignore development at their peril. Along the way, we revisit the main assumptions underlying the analogy as relational priming (ARP) account of analogy, clarifying and elaborating as necessary.

The understanding of decision-making systems has come together in recent years to form a unified theory of decision-making in the mammalian brain as arising from multiple, interacting systems (a planning system, a habit system, and a situation-recognition system). This unified decision-making system has multiple potential access points through which it can be driven to make maladaptive choices, particularly choices that entail seeking of certain drugs or behaviors. We identify 10 key vulnerabilities in the system: (1) moving away from homeostasis, (2) changing allostatic set points, (3) euphorigenic “reward-like” signals, (4) overvaluation in the planning system, (5) incorrect search of situation-action-outcome relationships, (6) misclassification of situations, (7) overvaluation in the habit system, (8) a mismatch in the balance of the two decision systems, (9) over-fast discounting processes, and (10) changed learning rates. These vulnerabilities provide a taxonomy of potential problems with decision-making systems. Although each vulnerability can drive an agent to return to the addictive choice, each vulnerability also implies a characteristic symptomology. Different drugs, different behaviors, and different individuals are likely to access different vulnerabilities. This has implications for an individual's susceptibility to addiction and the transition to addiction, for the potential for relapse, and for the potential for treatment.

The unified framework for addiction (UFA) formulated by Redish et al. is a tour de force. It uniquely predicts that there should be multiple addiction syndromes and pathways – a diversity that would reflect the complexity of the mammalian brain decision system. Here I explore some of the evolutionary and developmental ramifications of UFA and derive several new avenues for research.

The ten vulnerabilities discussed in the target article vary in their likelihood of producing temporary preference for addictive activities – which is the phenomenon that puzzles conventional motivational theory. Direct dopaminergic stimulation, but probably not the other vulnerabilities, may contribute to the necessary concavity of addicts' delay discounting curves, as may factors that the senior author analyzes elsewhere. Whatever their origins, these curves can themselves account for temporary preference, sudden craving, and the “automatic” habits discussed here.

Redish et al. suggest that their failures-in-decision-making framework for understanding addiction can also contribute to improving our understanding of a variety of psychiatric disorders. In the spirit of reflecting on the significance and scope of their research, I briefly develop the idea that their framework can also contribute to improving our understanding of the pervasive problem of procrastination.

Current computational models predict reward based solely on learning. Real motivation involves that but also more. Brain reward systems can dynamically generate incentive salience, by integrating prior learned values with even novel physiological states (e.g., natural appetites; drug-induced mesolimbic sensitization) to cause intense desires that were themselves never learned. We hope future computational models may capture this too.

Redish et al. provide a significant advance in our understanding of addiction by showing that the various addictive processes are in fact all decision-making processes and each may undergird addiction. We propose means for identifying more central addiction processes. This recognition of the complexity of addiction followed by identification of more central processes would help guide the development of prevention and treatment.

Redish et al. outline 10 vulnerabilities in the decision-making system that increase the risks of addiction. In this commentary I examine the potential role of social influence in exploiting at least one of these vulnerabilities, and argue that the needs satisfied by social interaction may play a role in decision-making with regard to substance use, increasing the risks of addiction.

Defining brain mechanisms that control and adapt motivated behavior will not only advance addiction treatment. It will help society see that addiction is a disease that erodes free will, rather than representing a free will that asks for or deserves consequences of drug-use choices. This science has important implications for understanding addiction's comorbidity in mental illness and reducing associated public health and criminal justice burdens.

The consideration of gambling as a decision-making disorder may fail to explain why the majority of people gamble, yet only a small percentage of people lose control of their behaviour to the point where their gambling becomes problematic. The application of dual process theories to gambling addiction offers a means of explaining the differences between “normal” and “problem” gambling, augmenting the multiple vulnerabilities proposed by Redish et al.

The framework for addiction offered by the target article can perhaps be simplified into fewer, more basic, vulnerabilities. “Impulsivity” covers a number of vulnerabilities, not just enhanced delay discounting. Real-world drug-use decisions involve both delay and probability discounting. The motivational salience of, and attentional bias for, drug cues may be related to a number of vulnerabilities. Interactions among vulnerabilities are of significance and complicate the application of this framework.

The “unified framework” for addiction proposed by Redish and colleagues is only unified at a reductionist level of analysis, the biological one relating to decision-making. Theories of addiction may be complementary rather than mutually exclusive, suggesting that limitations of individual theories might be unified through the combination of ideas from different biopsychosocial “complex” systems perspectives.

Although Redish et al. have pulled together a large number of approaches to understanding decision-making and common errors in cognition, they have outlined neither the necessary nor the sufficient attributes of addiction. They are correct in claiming that addiction is multifaceted and probably more akin to a syndrome than a genuine disease. But grasping what that multifaceted syndrome is still eludes us.

We commend Redish et al. for the progress they have made in bringing a measure of theoretical order to the processes that underlie drug addiction. However, incorporating information about situations in which drug users do not exhibit faulty decision-making into the theory would greatly enhance its generality and practical value. This commentary draws attention to the relevant human substance abuse literature.

Redish et al. view addictions as errors arising from the weak access points of the system of decision-making. They do not analytically distinguish between addictions, on the one hand, and errors highlighted by behavioural decision theory, such as over-confidence, representativeness heuristics, conjunction fallacy, and so on, on the other. Redish et al.'s decision-making framework may not be comprehensive enough to capture addictions.