Conditioned Inhibition and its Relationship to Impulsivity: Empirical and Theoretical Considerations.
Impulsivity is a widely used term in many fields of behavioral science, such as personality psychology, behavioral ecology, psychiatry, behavior analysis, cognitive neuroscience, behavioral economics, learning theory, and psychopharmacology (Green & Myerson, 2013). This term has been usually considered to be opposed to self-control, both terms being on opposite extremes of a continuum (e.g., Logue, 1995). Impulsivity is typically viewed as a multifactorial construct that includes reduced sensitivity to negative consequences (i.e., poor response inhibition), reduced sensitivity to delayed consequences, both rewards and punishments (i.e., delay discounting), acting without considering available information (i.e., inattention), willingness to seek novel experiences (i.e., novelty seeking), and engaging in risky behaviors (i.e., risk proneness; Evenden, 1999). In addition, impulsivity is also invoked as a core factor in several psychiatric conditions, such as attention deficit hyperactivity disorder (ADHD) and substance abuse (American Psychiatric Association, 2013). It is important to note that impulsivity must be considered a relative concept; that is, one can use the label "impulsive" to describe someone whose behaviors conform to one or more of the abovementioned descriptions, but only if this tendency is significantly greater than an to "individual baseline (i.e., state impulsivity) or a population of reference (i.e., trait impulsivity; Killeen, 2015; Odum & Bauman, 2010).
There are several procedures to assess the impulsivity/self-control dimension (hereafter "impulsivity") in laboratory settings. One of the advantages of such standard laboratory protocols is that they enable researchers to study the same phenomena in human and nonhuman organisms (Richards, Gancarz, & Hawk, 2011; Winstanley, 2011). Each of these procedures is supposed to evaluate different aspects of impulsivity. They can be classified into three broad categories: (1) choice, (2) omission, and (3) reaction-time procedures. Choice procedures are designed to assess whether an individual discounts the value of a reward (or of an aversive consequence; see Rodriguez, Bouzas, & Orduna, 2017) across increasing delays or decreasing probabilities of receiving that reward, thus measuring sensitivity to delayed outcomes and risk proneness, respectively (Rachlin, Rainieri, & Cross, 1991). In turn, omission procedures, such as differential-reinforcement of low-rates schedules (e.g., Stewart et al., 2006), stop-signal procedures (Liu, Heitz, & Bradberry, 2009), and go/no-go procedures (e.g., Trommer, Hoeppner, Lorber, & Armstrong, 1988), are supposed to evaluate an individual's capacity to withhold or inhibit a response when it prevents the delivery of a reward (i.e., maladaptive). Lastly, reaction-time procedures are assumed to measure attentional processes (e.g., Navarra, Comery, Graf, Rosensweig-Lipson, & Day, 2007).
Regarding the multiple factors comprising the constaict of impulsivity, some authors claim that different procedures in fact measure analogous behavioral tendencies that are only superficially different (e.g., Logue, 1988; Monterosso & Ainslie, 1999; Sosa & dos Santos, 2018; Tomie, Aguado, Pohorecky, & Benjamin, 1998). According to these authors, such superficial differences between procedures yield the wrong postulate of different underlying processes. The main idea behind these attempts at unification is that manifestations of impulsivity in a particular situation are expressions of a behavioral phenotype that is expressed in a similar fashion within analogue environmental constraints. This approach predicts a correlation among different measures of impulsivity (but see Green & Myerson, 2010). However, there seems to be little or no such correlations among the various measures of impulsivity (see Evenden, 1999; Lopez, Alba, & Orduna, 2017; Sonuga-Barke, 2005; Winstanley, Eagle, & Robbins, 2006; Winstanley, Dalley, Theobald, & Robbins, 2004), and only sparse evidence that supports it (Anker, Zlebnik, Gliddon, & Carroll, 2009; El Massioui et al., 2016; Tomie et al., 1998).
Some authors (e.g., Bari & Robins, 2013; Buss & Plomin, 1975) have proposed that a deficit in response inhibition is the core feature of impulsivity, whereas others (e.g., Richards et al., 2011) have argued that response inhibition is but one of the components of impulsivity (together with reduced sensitivity to delay, inattention, etc.). Although the capacity to inhibit a response has been the focus of much recent research in some fields, other fields (e.g., behavior analysis) have avoided the study of this construct, possibly due to a historical distrust of the term (see Hearst, Besley, & Farthing, 1970; MacLeod, Dodd, Sheard, Wilson, & Bibi, 2003). In this article, we will describe some of the most commonly used procedures to evaluate inhibition, their relationship to the concept of impulsivity, and their correlates with other measures of impulsivity. We focus on one of these procedures, conditioned inhibition, which has fallen into neglect in the impulsivity literature. We will review some of the reasons that discourage the use of this paradigm to study impulsivity, namely historical controversies, methodological hindrances, and the apparent mismatch between conditioned inhibition and traditional definitions of response inhibition. Finally, we contend that theoretical work on conditioned inhibition may provide insights into manifestations of impulsivity under different laboratory arrangements.
Procedures that Measure Response Inhibition
Response inhibition can be broadly defined as a process in which a particular response is stopped or prevented by the action of some agent, whereas the capacity for executing this response is retained and can be manifested as soon as the restraining agent is removed (adapted from Brunton, 1883, cited in Bari & Robins, 2013). In this section, we will review some of the most straightforward and/or most used paradigms to assess response inhibition. There are a number of other paradigms to assess this component of impulsivity, but they are beyond the scope of this manuscript. We roughly excluded paradigms that typically require more than one type of response (e.g., fixed consecutive number paradigm; see Bardo, Cain, & Bylica, 2006), evaluate more than just the response inhibition component of impulsivity (e.g., five-choice serial reaction time task and discrimination reversal learning; see Izquierdo et al., 2010; Robinson et al., 2009; Winstanley et al., 2003), or that are suitable mostly for humans (e.g., Stroop task; see Scheres et al., 2003; but see Haddon & Killcross, 2005, for a putative analogue procedure for rats).
Go/No-Go and Stop-Signal Tasks
The go/no-go task and the stop-signal task (SST) are designed to evaluate response inhibition and are extensively used to study impulsivity. Although they are procedurally similar and empirically correlated (see Reynolds, Ortengren, Richards, & De Wit, 2006), they are supposed to measure two dissociable processes. In the go/no-go task, subjects are exposed to separate "go" and "no-go" trials, each typically signaled by similar, but distinguishable, exteroceptive stimuli. Subjects are required to emit a specified response to obtain a reward in "go" trials, whereas they are required to either omit this response or perform a specific alternative response in "nogo" trials, or else they lose the reward. The SST consists of roughly the same conditions, but the "go-signal" is presented in all trials and the "stop-signal" is unexpectedly presented following a certain period of time from the onset of the "gosignal" in some trials (Bari & Robins, 2013). If the subject fails to refrain from responding in a stop trial, the time elapsed between the "go-signal" and the presentation of the "stop-signal" is decreased for the next trial. Conversely, if the subject successfully omits responding, this interval is increased. By means of this adjustment, researchers are able to estimate the interval at which subjects omit responding in 50% of the stop trials, which could be used as an index of response inhibition (Richards et al., 2011). The assumed difference between these procedures is that the go/no-go task measures the ability to withhold an inappropriate response before it arises, whereas the SST measures the ability to stop or cancel an already initiated response, in order to get a reward (or avoid a reward loss).
Some studies have evaluated whether performance in go/ no-go and stop-signal procedures is related to behavioral measures used to assess components of impulsivity apart from inhibition deficit. For example, Reynolds et al. (2006), Stahl et al. (2014), and Mackillop et al. (2016) found that performance in neither the go/no-go nor the SST correlates with performance in a procedure that measured delay discounting in humans. In contrast, using rats as subjects, Anker et al. (2009) found a relation between performance in a go/no go procedure and performance in a delay discounting procedure. Subjects were first trained in an intertemporal choice procedure (involving a trade-off between an immediate reward and a larger more delayed reward), and then were classified as "low-impulsivity" or "high-impulsivity" rats depending on their performance. Afterward, they were exposed to a go/nogo procedure using saccharin pellets or intravenous cocaine infusions as reward. High-impulsivity rats made more commission errors (i.e., emitted more responses during the "nogo" signal) compared with "low-impulsivity" rats when saccharin pellets were used as reward. This result suggests a common underlying process controlling performance on the two different tasks. However, there were no significant group differences in go/no-go performance when cocaine infusions were used as reward. This latter finding was interpreted by the authors as the result of a ceiling effect.
There is evidence of correlation between SST performance and self-report measures of impulsivity (e.g., Logan, Schachar, & Tannock, 1997; Vigil-Colet & Codorniu-Raga, 2004). However, several studies have found an absence of correlation between these measures and self-report impulsivity questionnaires (Enticott, Ogloff, & Bradshaw, 2008; Reynolds et al., 2006; Wilbertz et al., 2014). This absence of correlation has been interpreted as evidence of the multifactorial nature of impulsivity (e.g., Nombela, Rittman, Robbins, & Rowe, 2014), but another plausible interpretation has to do, instead, with the validity of self-report questionnaires. Bari and Robins (2013) suggest that there are considerable problems with self-report measures; for example, they could be biased by participants' past life events and self-perceptions, and they use "lexical categories that may have different meanings for different subjects" (p. 53). Yet, some studies have shown that other measures, like delay discounting (Solanto et al., 2001) or IQ (Horn, Dolan, Elliott, Deakin, & Woodruff, 2003) are better predictors of self-reported questionnaires' scores than response inhibition measures. Perhaps, the greatest strength of these paradigms is that they do correlate with maladaptive behaviors classified as reflective of pathological impulsiveness, such as antisocial personality disorder (Swann, Lijffijt, Lane, Steinberg, & Moeller, 2009), alcohol dependence (Lawrence, Luty, Bogdan, Sahakian, & Clark, 2009), disruptive behavior (Dougherty et al, 2003), eating disorders (Rosval et al., 2006), psychotic behavior (Huddy et al., 2013), and ADHD (Iaboni, Douglas, & Baker, 1995; but see Alderson, Rapport, Sarver, & Kofler, 2008).
Differential Reinforcement of Low Rates
Differential reinforcement of low rates (DRL) is a free operant procedure in which subjects must pace their responses in order to obtain a reward. In this situation, reward is contingent upon a response only if that response is spaced a specific minimum time from the previous emitted response (Kramer & Rilling, 1970). Premature responses reset the clock so that individuals must wait for the next response to be rewarded, whereas waiting too long to respond inevitably reduces reward rate. In order to maximize rate of reward, subjects must accurately measure time and inhibit any premature response (Richards, Sabol, & Seiden, 1993). There is evidence that impulsive humans and nonhuman animals fail to perform optimally in this procedure. For example, in a study conducted by Orduna, Valencia-Torres, and Bouzas (2009), spontaneously hypertensive rats (SHR), an animal model of ADHD, were considerably inefficient (i.e., made more premature responses) compared with two other nonimpulsive strains of rats, in a DRL procedure; however, this difference faded with sufficient training. Likewise, van den Broek, Bradshaw, and Szabadi (1987) selected impulsive and nonimpulsive humans on the basis of their scores on a behavioral test (Matching Familiar Figures Test) and later evaluated their performance on a DRL procedure. These authors found that impulsive participants showed a greater proportion of premature responses, thus earning fewer monetary rewards.
Despite the apparent usefulness of the DRL procedure for evaluating inhibitory capacity, performance in this procedure may be influenced by factors other than response inhibition (e.g., time estimation, motivational, and motor factors), thus obscuring the role of inhibition in maximization of reward. To overcome this problem, some authors have proposed analytical techniques (Richards et al., 1993) or supplementary procedures (Sanabria & Killeen, 2008; Watterson, Mazur, & Sanabria, 2015) to dissociate the contribution of different processes to DRL performance.
The Case for Conditioned Inhibition
Conditioned inhibition is a paradigm in which subjects learn that a stimulus signals the absence of an outcome that would otherwise occur. Thus, an individual is exposed to presentations of a nonreinforced stimulus occurring in close proximity to another stimulus (either discrete stimulus or a contextual cue) that is otherwise paired with a biologically significant event (i.e., unconditioned stimulus or reinforcer; Savastano, Cole, Barnett, & Miller, 1999). There are several ways to implement this operation, but perhaps the most prototypical is Pavlovian conditioned inhibition, also known as feature-negative discrimination. In this procedure, two types of trials are presented in a random order: ones in which a conditioned stimulus (excitatory CS) is paired with an unconditioned stimulus (US), and others in which the same conditioned stimulus is presented in compound with another stimulus (inhibitory CS) but the US is omitted (Pavlov, 1927). This procedure is also termed "the A+/AX- procedure," where "A" stands for the excitatory conditioned stimulus, "X" stands for the inhibitory conditioned stimulus, "+" stands for the presentation of the unconditioned stimulus, "-" stands for the omission of the US, and "/" denotes that the two types of trial are interspersed in the same situation.
Conditions for Inducing Conditioned Inhibition
Procedures known to induce conditioned inhibition entail the so-called expectancy violation, which can be defined as the surprising nonoccurrence of a signaled consequence (Williams, 1996). Expectancy violation can be readily identified in feature-negative discrimination procedures, in which nonreinforced trials (i.e., AX-) include a strong source of conditioned excitation (stimulus A) that is paired with US presentations. In differential conditioning (A+/X-) and explicitly unpaired (+/X-) procedures, however, there is no discrete excitatory stimulus accompanying nonreinforced trials. Rather, what is thought to produce expectancy violation in these cases is the presence of presumably excitatory contextual cues (Wagner & Rescorla, 1972). Following this rationale, the more salient the concurrent excitatory stimulation, the greater inhibitory power the target stimulus would acquire. Indeed, there is evidence that supports this idea (see Papini & Bitterman, 1993; but see LoLordo & Fairless, 1985). Another feature of most conditioned-inhibition procedures is a negative contingency (e.g., Rescorla, 1969a) between a cue (CS) and an outcome (mostly an US, but sometimes another CS; for instance, in sensory preconditioning and second-order conditioning). That is, the probability of US given a presentation of the CS, p(US|CS), must be lower than the probability of US given the absence of the CS, p(US|[logical not]CS). The value of the contingency between the CS and the US can be calculated as follows:
[DELTA]p = p(US|CS) - p(US|[logical not]CS)
Where [DELTA]p is the value of contingency or statistical relationship between the CS and the US, p(US|CS) is calculated as the frequency of occurrence of the CS and the US together (in a given fraction of time) divided by the total frequency of the CS, and p(US|[logical not]CS) is calculated as the frequency of occurrence of the US in absence of the CS divided by the total frequency of the US (Williams, 1996). Assuming that different trials occur with equal frequency, [DELTA]p for stimulus A in differential conditioning (A+/X-) is 1.0, whereas it is 0.5 for the same stimulus in feature-negative discrimination (A+/ AX-). Conversely, the value of [DELTA]p for stimulus X in both procedures equals -1.0. There is evidence that the amount of conditioned inhibition is proportional to the degree of negative contingency between the target stimulus and the US (Rescorla, 1969a; but see Hearst & Franklin, 1977).
Furthermore, in conditioned-inhibition protocols, the target stimulus not only signals the omission of the US, but it also provides information about when the US will occur (Gallistel & Gibbon, 2000). Thus, individuals exposed to this procedure could leam that when the target stimulus is presented the US will not occur within a specific lapse of time, namely the programmed minimum intertrial interval, which usually is relatively long. We will refer to this attribute as time-out signaling. It has been asserted that inhibitory properties develop in a particular stimulus when it signals a long period without the occurrence of the US. For example, if one alternates A+ and X- trials using a fixed time 60-s schedule, then the stimulus X signals a reliable 60-s period without the US befalling. Alternatively, if one uses a variable time 60-s schedule with a minimum value of 20 s, then the stimulus X signals a period of time of at least 20-s without the US happening. There is some evidence supporting that this temporal signaling contributes to the development of inhibitory properties in a stimulus independently of other factors (e.g., Moscovitch & LoLordo, 1968).
Tests for Detecting Conditioned Inhibition
There are multiple ways to assess the inhibitory properties of a stimulus, each related to what is thought to be learned in conditioned-inhibition procedures. For instance, individuals may learn to perform a response tendency opposed to that elicited by the excitatory stimulus (Rescorla, 1969b). In line with this assertion, Rescorla (1969b) recommended two basic tests for conditioned inhibition, summation and retardation, that must be applied together in order to rule out alternative explanations regarding attentional shifts (i.e., increased or decreased orientation to the target stimuli that may produce misleading changes in behavior). The summation test consists of presenting the putative conditioned inhibitor in a novel compound with a conditioned excitor. If the strength of the conditioned response elicited by this compound is less than the one elicited by the excitor alone, it is assumed that the putative inhibitor antagonized conditioned responding to some extent. In order to exclude the possibility that any such difference was due to the inhibitor drawing the attention required for the excitor to elicit responding, a retardation test is also required. The retardation test involves pairing the putative inhibitory stimulus with the US. If summation test results were due to attention being misdirected to the inhibitor, then its pairings with a US would result in fast acquisition because it already attracts attention. In contrast, if the development of conditioned responding by the target stimulus is relatively slow one may assume that the summation test results were due to inhibition.
The assumption that inhibition is an opposite process to excitation has the additional implication that inhibitory conditioning with a US of an affective class (appetitive or aversive) would endow a CS with an opposite affective value (LoLordo & Fairless, 1985). This means that an inhibitor of an appetitive US would function as an aversive conditioned stimulus and, conversely, an inhibitor of an aversive US would function as an appetitive conditioned stimulus or conditioned reward (Savastano et al., 1999). For example, in a study conducted by Wasserman, Franklin, and Hearst (1974), pigeons were exposed to a condition in which a specific light color projected on a key predicted food and the projection of a different color predicted the absence of food. Pigeons were observed to approach and peck the key when the light associated with food was on, which is known as autoshaping or sign-tracking behavior. Sign-tracking behavior is considered as a reliable measure of how much incentive motivation has been acquired by the cue (Anselme, 2015). It is crucial to note that the birds were also observed to recede from the key when it was illuminated by the light color associated with the absence of food (i.e., inhibitory cue). This could be regarded as evidence of the aversive properties acquired by the inhibitory cue. In addition, there is empirical support for the converse: namely, the acquisition of appetitive properties by conditioned inhibitors in an aversive situation (e.g., Fernando, Urcelay, Mar, Dickinson, & Robbins, 2014; LoLordo, 1969).
Finally, an old method for assessing conditioned inhibition that has fallen into disuse is the inhibitory generalization gradient (Terrace, 1966). In this procedure responding to a putative conditioned inhibitor is evaluated by varying this stimulus along some physical dimension (e.g., color, pitch, tilt). If the lowest response strength is observed to the trained stimulus and the response strength increases as the physical dimension is varied, then the trained stimulus is thought to possess inhibitory properties (Hearst et al., 1970). A sharp "U" shaped gradient would be indicative of strong inhibitory properties, whereas a flat function would be indicative of mild or no conditioned inhibition acquired by the stimulus.
Conditioned Inhibition and Impulsivity
Conditioned-inhibition procedures are notably similar to SST and go/no-go procedures in that subjects are typically exposed to two types of trials, those in which responding is "appropriate" and those in which responding is "inappropriate." The key difference between these two subcategories of procedures is that SST and go/no-go paradigms impose an instrumental or operant contingency (i.e., response-outcome) on responding in the "inappropriate" trials. On the other hand, in conditioned-inhibition procedures the imposed contingency is of Pavlovian nature (i.e., stimulus-stimulus); simply put, the outcome is independent of responding. Whether this difference is crucial to the validity of such procedures as measures of impulsivity is a question that only can be answered empirically. However, there are no studies systematically comparing these procedures.
There is scant work relating impulsivity and conditioned inhibition. For example, He, Cassaday, Howard, Khalifa, and Bonardi (2011) assessed conditioned inhibition in participants with a history of violent and sexual offences classified as "impulsive." In this study, conditioned inhibition was induced by presenting pictures of Lego blocks (CS) paired with either positive (excitatory trials) or neutral images (inhibitory trials). The critical trials, AB+ and AX-, were intermixed with other types of control trials. Conditioning was measured by asking participants to predict what kind of picture would follow the presentation of a given compound CS. Conditioned inhibition was assessed via a summation test, in which it was observed whether the inhibitory properties of the stimulus X transferred to a condition different from that of the training phase. He et al. found that offenders showed an impaired performance relative to control participants. However, members of the same research team also found no correlation between performance in a similar task and scores of self-report questionnaires of impulsivity in graduate students (He, Cassaday, Bonardi, & Bibby, 2013). These results, again, raise doubts about the validity of either self-report or conditioned-inhibition measures.
To our knowledge, there are just two studies relating conditioned inhibition and impulsivity in nonhuman subjects. Bucci, Hopkins, Keene, Sharma, and Orr (2008) compared performance in a Pavlovian conditioned-inhibition protocol of male and female rats of two different strains in an appetitive procedure using approach responses to the feeder as an index of conditioned responding. These authors used the SHR strain as well as Wistar Kyoto (WKY) rats, a usual control strain for SHR. Besides the expected differences between these two strains, they found an impaired development of discrimination between A+ trials and AX- trials in the SHR strain, but only for females. Bucci et al. (2008) concluded that their results resemble clinical data, in which females diagnosed with ADHD display more extreme cognitive deficiencies (e.g., Gershon, 2002), which, according to them, validates this procedure as a model of impulsivity. In another study, Green, Chess, Conquest, and Yegla (2011) compared the performance in tests of conditioned inhibition of SHR and Wistar rat strains after training them on a Pavlovian conditioned-inhibition protocol using an eyeblink conditioning procedure. These authors did not find significant differences between strains in the level of conditioned inhibition as measured by either the summation or the retardation test. Green et al. (2011) interpreted this result appealing to neurophysiology: unlike other tasks, inhibition of the eyeblink does not involve the medial prefrontal cortex, which is known to have impulsiveness-related abnormalities in the SHR strain (e.g., Russell, de Villiers, Sagvolden, Lamm, & Taljaard, 1995).
Possible Causes of the Neglect of Conditioned Inhibition in the Field of Impulsivity Research
Given the aforementioned similarities between conditioned-inhibition procedures and other protocols of impulsive behavior (i.e., go/no-go and SST), the scarcity of research on impulsivity using conditioned-inhibition paradigms is noteworthy. Indeed, the disregard of this protocol is evident not only from the number of empirical papers available, but also from the little theoretical work considering the possibility of a relationship between these fields. For example, in their recent extensive review of the relationship between impulsivity and inhibition, Bari and Robins (2013) barely mention the conditioned-inhibition paradigm, and they do so only as an incidental early approach to the study of inhibition proposed by Pavlov (1927). In this section, we argue that such disregard may be explained, in part, as a product of a historical controversy around conditioned inhibition, much of which has now abated. Then, we suggest that another factor that may discourage the study of conditioned inhibition in the field of impulsivity research is the perceived necessity of the two tests advocated by Rescorla (1969b). We also discuss a third possible explanation for the disregard of this paradigm: its apparent mismatch with traditional definitions of response inhibition. Afterward, we elaborate how the theoretical work around conditioned inhibition can be applied to understand performance in some tasks that measure impulsive behavior.
Criticism of Conditioned Inhibition from Learning Theorists
Some influential authors have expressed theoretical and empirical concerns about conditioned inhibition. Those concerns may be classified as concerns with inhibition itself as a construct, concerns with the nature of inhibition as a process, and concerns with the necessary and sufficient conditions to demonstrate inhibition. One of the major detractors of this concept was Skinner (1938), who expressed concerns about its logical validity, ontological status, and how it compromises parsimony. For example, appealing to parsimony, he argued that the organisms simply learn to respond in conditions (taking into account the overall set of stimuli present) in which they experience reinforcement and learn not to respond in conditions in which they do not experience reinforcement. Therefore, according to this author, there is no need for an additional construct accounting for the observed absence of responding in a particular case. Another concern is that conditioned inhibition theorizing was originally inspired by Pavlov's (1927) intuitions about its underlying neural mechanisms, whereas his study of this phenomenon was purely behavioral. Skinner (1938) warned that this kind of extrapolations might potentially obfuscate theorists, as they may lose track of what they know and what they ignore (see Zilio, 2016).
Other authors, like Donahoe and Palmer (1988), highlighted that the main procedures used to measure acquired inhibitory properties in a stimulus rely on inferring its influence on an excitatory process. According to these authors, this inference would be valid only if all other variables that could influence excitatory processes are held constant. They added that the independence of inhibition from these confounding factors cannot be fully ruled out without direct measurement of the former. In a thorough review, Donahoe and Palmer (1988) applied this reasoning to summation, retardation, and generalization gradient techniques, but they conveniently omitted the approach and withdrawal test for conditioned inhibition (e.g., Wasserman et al., 1974). This protocol can be considered a demonstration of inhibition as a behavior tendency opposed to excitation without the need for further tests, given that the withdrawal (a behavior tendency opposed to approach, which is the index of excitatory conditioning) from the focalized inhibitory CS can be observed at the time of training.
Another example of direct measuring of conditioned inhibition comes from a study conducted by Tobler, Dickinson, and Schulz (2003). In this study, macaques were trained in a feature-negative discrimination (A+/AX-) paradigm with an appetitive unconditioned stimulus. Activations above baseline responding of single dopamine neurons in the midbrain were recorded with intracranial electrodes as the conditioned response. They found that the X stimulus not only counteracted the activations produced by a conditioned excitor (i.e., summation), but it also depressed the dopamine neurons activity below the baseline when presented alone. This also qualifies as direct evidence of the effects of a conditioned inhibitor. In sum, there was evidence of direct measurement of conditioned inhibition back in the time when Donahoe and Palmer (1988) wrote their paper (Wasserman et al., 1974; see also Batson & Best, 1981), and there is also recent evidence showing that conditioned inhibition can be directly measured.
In another controversial paper, Papini and Bitterman (1993) likewise disputed the necessary and sufficient conditions for demonstrating conditioned inhibition. These authors reviewed the empirical literature about conditioned inhibition and concluded that there were few studies (if any) that have unambiguously demonstrated this phenomenon. They urged researchers to conduct studies with full counterbalancing of crucial stimuli and in which tests for conditioned inhibition were performed in equality of conditions for independent groups. In addition, they recommended some control conditions that would serve to rule out alternative explanations. To date, there are some demonstrations of conditioned inhibition that have followed Papini and Bitterman's (1993) recommendations, for example, Cole, Bamet, and Miller (1997) using rats as subjects in an aversive conditioning paradigm, and Acebes, Solar, Moris, and Loy (2012) with snails as subjects using an appetitive paradigm.
Although several reviews of conditioned-inhibition literature have highlighted all of the above cited (and even more) theoretical and methodological issues in the study of this topic, some of them have also been optimistic regarding the use of the concept and further theorizing and researching about it. For example, Hearst et al. (1970) stated that,
If Pavlov had been a little less excessive in his neurological speculations, and Skinner a little less excessive in his demands for parsimony, the experimental analysis of inhibitory effects might be a more popular area of... research today. There seems no good reason why neglect of the topic should continue, (p. 406)
We strongly agree with Hearst et al. (1970) and we consider the study of conditioned inhibition to be fertile ground for theorizing about the learning processes involved in it. However, it would be a challenge to set forth research designs qualified to answer current and new questions about inhibition. As Savastano et al. (1999) declared,
... one could argue that behavior indicative of inhibition is complex, and multiple theoretical constructs may be necessary to account for behavioral inhibition across many situations [but] to reject the original [appealingly simplistic] view of inhibition as the opposite of excitation does not entirely diminish the utility of inhibition as a theoretical construct in animal learning, (pp. 120-122)
Another example is the review by Williams, Overmier, and LoLordo (1992) in which the authors concluded that "there are no new theories or data that seriously undermine the necessity of the construct of inhibition" (p. 287). In summary, no matter how harsh the concerns about conditioned inhibition, some authors tend to favor the explicative and/or heuristic utility of the concept. We would like to subscribe to this spirit and encourage further development in the field which, it is important to note, might afford promising linkages with other fields and approaches. We suspect that the controversies described in this subsection considerably discouraged the use of conditioned-inhibition paradigm in some fields and that are partly responsible for the neglect of conditioned inhibition as a paradigm to study of impulsivity.
Difficulties of the Tests to Demonstrate Conditioned Inhibition
As has been noted by some authors (e.g., Papini & Bitterman, 1993; Williams et al., 1992), Rescorla's (1969b) recommendation of conjointly applying summation and retardation tests in order to unambiguously identify a stimulus as a conditioned inhibitor was taken too seriously. For example, Williams et al. (1992) stated that "most researchers, following Rescorla, either used both the summation and retardation tests or felt they should have" (p. 275). Including both tests of conditioned inhibition in a study considerably increases the amount of time, effort, and resources that should be devoted, which may constitute a technical barrier for some researchers. This is especially troublesome for studies that involve brain lesions or administration of drags, in which the several usual control conditions (e.g., sham lesions, vehicle administration) would have to be multiplied by two while keeping a sufficient group size. In addition, a conceptual hindrance may emerge from the fact that each test is assumed to assess different independent underlying processes (Miller & Matzel, 1988). This may constitute a potential issue at the time of justifying whichever of the tests is being selected as a measure representing the construct of conditioned inhibition. Moreover, performance in the summation and/or retardation reflects transitional states, whereas the most common practice is to measure variables that reach a steady state whenever one wants to compare means between groups (or within subjects) or for abstracting behavioral scores to be correlated with other variables (Killeen, 1978). All of these disadvantages may have discouraged the adoption of conditioned inhibition as a model of impulsivity.
An example of such underemphasis on conditioned inhibition can be seen in a different research area: associative learning models of anxiety disorders. The central concept in this field is safety signal, which can be defined as a stimulus that predicts the offset, absence, or reduction of a feared (i.e., aversive) outcome (Lohr, Olatunji, & Sawchuk, 2007). It is interesting that safety signals have the property of counteracting the effects of fear signals when they are presented together (e.g., Christianson et al., 2008) in a manner resembling the observed effect of conditioned inhibitors in summation tests. It is assumed that the process responsible for the development of safety signal properties in a cue is impaired in people with anxiety disorders (e.g., Haddad, Pritchett, Lissek, & Lau, 2012). In spite of the clear resemblance of safety signals to conditioned inhibitors, regarding both the conditions to induce them and those to test their effects, the former are often not presented as related to the latter (e.g., Lohr et al., 2007). This could be partly due to subtle differences in the methodological details used by each field, including whether or not to conjointly apply summation and retardation tests.
In their thorough reviews about the pertinence of the twotest strategy proposed by Rescorla (1969b), Williams et al. (1992) and Papini and Bitterman (1993) suggested that this rule of thumb is not as unerring as some may have assumed. Failing either the summation or the retardation tests should not always be considered as evidence of the absence of conditioned inhibition. Williams et al. (1992) asserted that in cases in which a stimulus possesses both inhibitory and excitatory properties (which has been found in some studies, e.g., Droungas & LoLordo, 1994), such stimulus could pass the summation but not the retardation test yet still should be considered a conditioned inhibitor. For their part, Papini and Bitterman (1993) likewise disputed the need for application of both tests in order to assess whether a stimulus has conditioned-inhibitory properties. They argued that "In ... several. .. cases, inhibition could be reasonably inferred from the results of a single test" (p. 348). This claim is founded on the idea that the attentional shifts that those tests are designed to rule out can be discarded early in the conditioning stage in some cases. Papini and Bitterman (1993) invoked the feature-negative discrimination paradigm (A+/AX-) as an example of this. In such a paradigm subjects usually respond substantially in A+ trials and relatively less in AX- trials. The simple fact that subjects respond differentially in the presence of the stimulus X appears to be sufficient evidence to suggest that they are attending to it. Therefore, a relatively slow acquisition of the conditioned response to the X stimulus (i.e., retardation) following a successful A+/AX- training could not be attributed to a decreased attention to the target stimulus. A reasonable interpretation in this case would be that the training phase has endowed the stimulus X with inhibitory properties.
Aside from the controversy of whether or not to conjointly apply the two-test strategy, it should be noted that this protocol was originally conceived by Rescorla (1969b) as a mean of demonstrating that a given stimulus have acquired conditioned-inhibitory properties under certain circumstances. If one's study aims to inquire about a putative inhibitory phenomenon it may be valid to use a procedure that previous studies have reasonably demonstrated give rise conditioned inhibition, even if one's study does not attain the exact methodological rigor.
Indeed, some contemporary papers have ventured to study conditioned inhibition without using all the required controls and yet may be considered to contribute to the field with valuable data. For example, Harris, Kwok, and Andrew (2014), using an adapted version of the A+/AX- procedure in an appetitive situation with rats, studied the conditions needed for a stimulus that signals a reduction in the expected rate of US to exhibit inhibitory properties. In this study, conditioned inhibition was assessed only by summation tests. It is interesting that some evidence within this study suggested that retardation tests were not needed. Harris et al. (2014) consistently observed that the effectiveness of the inhibitory stimulus was greater when presented in compound with a strong (associated with a relatively high US rate) excitatory CS in the summation test. As previously mentioned, the purpose of the retardation test is to rule out that the decreased responding to the excitatory CS in the summation test is not due to an attentional shift toward the supposed inhibitory stimulus. One would reasonably expect that an intrusive stimulus would more readily outcompete a relatively weak CS for response control. Therefore, it can be assumed that the suppressing effect observed in the summation tests from Harris et al.'s (2014) study was likely due to conditioned inhibition.
Studies in which it is claimed that conditioned inhibition is being assessed but in which neither summation nor retardation tests are used represent more extreme cases. For example, in the studies conducted by MacLeod, Potter, Simoni, and Bucci (2006) and by Meyer and Bucci (2014a) the selected measure of conditioned inhibition was the degree of discrimination in an A+/AX- procedure. According to Papini and Bitterman (1993), "[ljess responding to [AX] than to A does not demonstrate that [X] is inhibitory, but suggests only the hypothesis that [X] is inhibitory, which must [emphasis added] be evaluated independently of training" (p. 348; see also Stout, Escobar, & Miller, 2004). Nevertheless, this may depend on the definition of inhibition that is adopted. If one uses the definition that we provided previously (a process in which a particular response is stopped or prevented by the action of some agent) there seems to be no problem with considering discrimination in an A+/AX- procedure as an instance of inhibition.
Does Lack of Conditioned Inhibition Have Negative Consequences?
A typical definition of response inhibition deficit includes (1) a component of failure in restraining or stopping a response, and (2) a component regarding the reduced sensitivity of that particular response to negative (i.e., adverse) consequences or its situational maladaptiveness (see, e.g., Bari & Robins, 2013; Richards et al., 2011). The first component can be easily identified in DRL, go/no-go, SST, and conditioned-inhibition performance. Nevertheless, the second component can be identified in the former three procedures but not so clearly in the conditioned-inhibition paradigm. As we explained earlier, in DRL, go/no-go, and SST procedures there are explicit operant contingencies arranged so that the failure to restrain the target response leads to actual losses of reward. However, strictly speaking there are no adverse consequences for an individual that fails to restrain conditioned responses in a conditioned-inhibition paradigm. Furthermore, it would be hard to assert that, within a laboratory setting, responding during an inhibitory stimulus is maladaptive. We believe that this apparent feature of the conditioned-inhibition paradigm is another tentative foundation of the prevailing breach between this paradigm and the concept of impulsivity. In the following paragraphs, we provide some arguments to reduce this conceptual gap.
The first argument is that conditioned inhibition is a laboratory model of a learning phenomenon assumed to occur in natural settings. It is safe to suppose that in natural settings the failure to restrain a response in anticipation of an outcome that is not going to happen could entail risks for any organism. For example, approaching a cue usually related to reward when it is accompanied by a cue that warrants that the reward will not be available on this occasion could leave an animal vulnerable to a predator's attack. In less extreme cases, such response could also lead the organism to lower its guard in defending its territory from a conspecific, or even distract it from actually obtaining a reward. Then, conditioned inhibition could be contemplated not as a measure of the learned capacity to passively avoid reward loss (like the other inhibition paradigms) but, it is interesting to note, as an index of a "built-in" or phylogenetically determined process that serves to elude potentially adverse consequences.
Although there are no explicit aversive contingencies for responding during the conditioned-inhibitory stimulus in the laboratory setting, every response, either musculoskeletal or endocrinal, carries an energy cost associated with it. This response cost can be equated with aversion (e.g., Rushen, 1986), derived from energy expenditure or even from the departure from other, preferred, behavioral pattern ("leisure"; Rachlin, Kagel, & Battalio, 1980). In fact, some authors have considered the minimization of costs an important factor for optimality of behavior (Alonso, Fairbank, & Mondragon, 2012). Following this reasoning, conditioned responses during the inhibitory stimulus should eventually be decreased by this "mild" operant contingency. Therefore, failure to restrain a response in the conditioned-inhibition paradigm can be actually seen as maladaptive.
One more thing to consider is the possible emotional/ affective effects of the surprising omission of an outcome. For instance, there is evidence that a signal for reward omission elicits, in addition to a cluster of autonomic responses (Papini & Dudley, 1997), a withdrawal from the area in which a reward is usually delivered (in opossums, Papini, 1988, and in rats Papini & White, 1994). This behavior can be considered as distinct from the one described by Wasserman et al., 1974) in which pigeons withdrew from the focalized signal for nonreward. In the experiments performed by Papini (1988) and Papini and White (1994) animals recede instead from the source of reward when the signal of nonreward was presented and not from the signal itself. The latter behavior was interpreted by Papini and White (1994) as analogous to escape responses from a frustrating situation. Of course, signals of the omission of reward would elicit an aversive state in the organism. Moreover, this emotional effect might be especially powerful near the area in which the animal gathers the food, perhaps due to the strong food anticipation prompted by those contextual cues. Hence, receding from the feeding area could be contemplated as having an implicit operant contingency; animals may be escaping from a state of discomfort. If that were the case, approaching responses during the presentation of an inhibitory stimulus can be considered as a failure to avoid an aversive consequence. This rationale appears to provide construct validity for conditioned-inhibition performance as a negative measure of impulsivity.
Although these arguments could be fairly regarded as conjectural, we recommend not overlooking them. The distance between conditioned inhibition and other procedures to measure inhibition may be just an artifact. We think that this gap is associated with the idea that there is a fundamental distinction between Pavlovian and operant behavior. A debate about this distinction is out of the scope of this article, but if there are well-known Pavlovian contingencies embedded in operant situations (Mowrer, 1956; Papini, 2001), the reverse assertion could also be tme. Operant contingencies may be implicit in Pavlovian paradigms more often than assumed (e.g., Holland, 1979). The next section is intended to elaborate on the former possibility in pointing at some of the aspects involved in procedures that induce conditioned inhibition in operant procedures to measure impulsivity.
Some Aspects Known to Induce Conditioned Inhibition Are Embedded in Paradigms of Impulsive Behavior
Discrete Trial Paradigms
In the go/no-go tasks, when "go" trials have a feature in common with "no-go" trials, the procedure resembles a feature-negative discrimination procedure (A+/AX-); and when "go" and "no-go" trials involve different discrete cues, the protocol resembles instead a differential conditioning procedure (A+/ X-). Both negative feature discrimination and differential conditioning procedures are known to induce conditioned inhibition (e.g., Cole et al., 1997; Miller, Hallam, Hong, & Dufore, 1991). As stated previously, the critical difference between conditioned inhibition and go/no-go tasks is that the latter involves an imposed operant contingency; namely, individuals are rewarded if they restrain from emitting the target response during the "no-go" trials. On the other hand, "nogo" trials in SSTs always involve a feature in common with "go" trials, which make them similar to feature-negative discrimination procedures. Nevertheless, the inhibitory cues in SSTs are usually presented following a time lapse from the excitatory cues and, it is important to note, this time lapse varies dynamically from trial to trial. This attribute, in addition to the imposed operant contingency, distinguish SSTs from conditioned-inhibition procedures.
The DRL Procedure
Performance in the DRL procedure is considered impulsive when subjects fail to space their responses (i.e., respond with a high frequency within a given period of time), leading to loss of reward. In such situations, responses are thought to be maintained by the reward delivery that is contingent upon them. Nonetheless, this contingency is not perfect, given that not all responses are followed by a reward; rather, only responses that satisfy the procedure's requirements yield reward delivery. For an optimal (or near optimal) performance, subjects must eventually discriminate the conditions in which their responses are followed by a reward. It has been suggested that stimulus discrimination in operant situations involves learning to inhibit nonreinforced responses (Bouton, Trask, & Carranza-Jasso, 2016). However, unlike traditional discrimination procedures, in DRL tasks typically do not involve external stimuli signaling when responding will be reinforced. Subjects that perform well in DRL procedures are said to have an accurate estimation of time (e.g., Sanabria & Killeen, 2008). It may be assumed that, in order to develop such timing accuracy, subjects have to be able to discriminate different interoceptive states and/or different behavioral patterns (Killeen & Fetterman, 1988; but see Donahoe & Burgos, 1999). Subjects might be in a distinctive behavioral state when a response was recently performed (hereafter referred as Stage 1) and when a proper amount of time for the next response to be rewarded has elapsed (Stage 2). Once stimuli occurring during Stage 1 acquire inhibitory properties, subjects may restrain from responding in their presence, thus avoiding reward loss.
But how would those stimuli acquire inhibitory properties? Such cues might become inhibitory through an expectancy violation phenomenon. Stimuli at Stage 1 are paired with cues that otherwise would signal the delivery of reward, like background stimuli and/or proprioceptive stimuli generated by approach and operation of the manipulandum. Those cues are always present nearby reward delivery and reward consumption, thus bearing considerable conditioned value. Hence, when responding and not getting a reward, subjects may experience an expectancy violation that could influence the value of stimuli present at that moment. Stimuli during Stage 1 also might acquire inhibitory properties given that target responses (i.e., lever pressing) in this stage have a negative contingency with the reward. Indeed, when scheduling events in DRL procedures one should impose a "negative punishment contingency" to responses emitted before the specified time criterion is accomplished; that is, a negative statistical relationship between the target response and reward delivery during Stage 1. Yet another way in which stimuli during Stage 1 could acquire inhibitory properties is that they signal a relatively long period to the next reward delivery. For example, in a 30-s DRL schedule subjects must space their responses at least 30-s in order to maximize available rewards. If an individual responds at second 15 after the last reinforced response, the timer resets and it must wait 30 more seconds to have the opportunity to obtain another reward. Thus, the cues in Stage 1, when presented in compound with response-related feedback, will always signal a period of 30 or more seconds without reward. Whenever the stimuli or behavioral states in the Stage 1 acquire inhibitory properties, the organism would tend to emit fewer premature responses, thus maximizing available rewards (which is indicative of self-control in this paradigm).
Intertemporal Choice Procedures
It is interesting that the conditions known to induce conditioned inhibition can also be found in other procedures to assess impulsivity not intended to evaluate the component of response inhibition. For instance, these features can be found in intertemporal choice procedures, which assess control by delayed outcomes. These procedures present subjects with a situation in which two reward alternatives are available: an immediate reward and a delayed but larger reward. Individuals that show a greater preference for the smaller sooner (SS) alternative are said to be more impulsive than the ones who prefer the larger later (LL) reward. It is important to note that choices of the SS alternative are followed by an intertrial interval adjusted so that the next choice opportunity happens at a time equivalent to that in which the individual would have chosen the LL alternative (Richards et al., 2011). Therefore, the only way to optimize the available resources is by choosing the LL alternative.
It is important to note that, unlike DRL procedures, in intertemporal choice procedures, the development of inhibitory properties to stimuli that control impulsive behavior cannot take part through the total omission of the reward, because the impulsive response does not cancel reward delivery but just reduces the overall reward gain in the long run. In this case, inhibitory properties may develop because of a stimulus association with a reduction in the expected reward. Indeed, there is evidence of acquisition of inhibitory properties by stimuli that signal the occurrence of a relatively low magnitude US (Mackintosh & Cotton, 1985; Wagner, Mazur, Donegan, & Pfautz, 1980) or a relatively low rate of US delivery (Harris et al., 2014). This can be considered as a partial omission of the US. If stimuli related to the SS alternative would acquire inhibitory properties, choices of this alternative would decrease, which may lead to a rise in preference for LL alternative, thus reducing impulsivity. There are studies showing some evidence that inhibitory learning can contribute to optimal behavior in intertemporal choice procedures (Sosa & dos Santos, 2014a, b) as well as in probabilistic choice procedures (Laude, Stagner, & Zentall, 2014; Trujano, Lopez, RojasLeguizamon, & Orduna, 2016). Unfortunately, these studies do not explain in detail how this inhibitory learning would take place.
Stimuli related to the SS alternative may acquire inhibitory properties through expectancy violation. Despite the conspicuous difference associated with each alternative (e.g., one vs. five food pellets), certainly the onset of SS reward delivery has stimulus elements in common with reward delivery of LL alternative (namely, sounds and vibrations within the experimental apparatus). Those stimuli could control strong conditioned responses in anticipation of the large reward. However, as soon as the subjects detect the partial omission of the large reward, they could display an emotional state of frustration that would be associated with the events that preceded this state (e.g., background and proprioceptive stimuli related to the operation of the manipulandum). Stimuli related to the SS alternative could also acquire inhibitory properties given that they have a negative contingency with the large reward associated with the LL alternative. Because SS and LL alternatives are related to mutually exclusive responses, the outcomes of each alternative should be associated with unique sources of response-related stimulation. Hence, stimuli related to the SS alternative would be negative predictors of the larger reward, in as much as they are never followed by it and the large reward is never preceded by them (i.e., [DELTA]p of-1.0).
One more way in which stimuli related to the SS alternative could acquire inhibitory properties is by signaling a considerable time amount before the individual can get the larger reward. Choices of the SS alternative are followed by a time-out from the opportunity to access the reward associated with the LL alternative. In order to compute the minimum time between an SS choice and the large reward, one should add the duration on the adjusted intertrial interval to the duration of the delay in the LL alternative. Stimuli related with a time out from reward are known to acquire aversive properties (see Leitenberg, 1965). Acquisition of aversive properties by stimuli related to the SS alternative would reduce the preference for that alternative.
The difference between intertemporal choice procedures and the available procedures to assess response inhibition may have discouraged efforts at developing conceptual bridges between them (but see Bari & Robins, 2013; Poulos, Parker, & Le, 1995). However, the differences between these types of procedures might not be as large as some authors have suggested. For example, Evenden (1999) pointed out two kinds of differences, those related to the underlying processes involved in each and those that have to do with the procedures themselves. This author suggested that performance in intertemporal choice procedures is affected by the subject's ability to assess subtle outcomes of their behavior, whereas the DRL procedure would be affected by the subject's organization of motor behavior. In what concerns to the details of each procedure, Evenden (1999) emphasized that in DRL procedures the subject's behavior during the waiting time has an important influence on whether or not the reward is delivered, whereas in intertemporal choice procedures "once the subject has made the [LL] choice the outcome is already decided, whatever it does during the delay" (p. 355). Therefore, it would be important to assess impulsivity by only varying this procedural feature and maintaining all other conditions the same, in order to scrutinize its impact on performance.
A study conducted by Reynolds, de Wit, and Richards (2002) sheds light on this question. They exposed a group of rats to a traditional intertemporal choice procedure (delay-discounting group) and another group of rats to a similar procedure but in which subjects could "defect" after choosing the LL alternative during the delay to reward period (delay-of-gratification group). This latter procedure is an analogue of the delay-of-gratification procedure, that was originally developed to study impulsiveness in children (e.g., Mischel, Shoda, & Rodriguez, 1989), which is assumed to share more features with response inhibition measures than intertemporal choice tasks (Reynolds & Schiffbauer, 2005). Reynolds et al. (2002) observed no significant differences in choice behavior between both groups of rats.
This result suggests that the apparently crucial feature of availability of the small reward during the waiting time is not a determinant for performance, at least for rats on these conditions. Reynolds et al. (2002) recorded the rate of defection responses in the delay-of-gratification group and compared it with those of the delay-discounting group, in which these responses had no programmed consequences. It is interesting that they found that the rate of defections was significantly lower on the delay-of-gratification group, showing that subjects were sensitive to the contingency associated with this response. More important, this result suggests that the outcome of defection responses (i.e., access to the smaller reward and all the associated stimuli) may have acquired aversive properties. Reynolds and Schiffbauer (2005) argued that these defection responses would decrease as being maladaptive because they prevent optimization after having invested time waiting for the larger reward. In our framework, this result could be accounted for by appealing to the expectancy violation, negative contingency, and signaling of a long period without reward, associated with defecting responses, similar to those described previously for SS alternative choices in intertemporal choice procedures.
Further Considerations of Relating Conditioned Inhibition to Impulsivity
Considerable work has been devoted to developing models that describe delay discounting, the process assumed to account for impulsivity in intertemporal choice procedures (see Madden & Bickel, 2010). In contrast, relatively little attention has been paid to the processes involved in the development of self-control. One of the proposed hypothetical processes for the development of self-control is soft-commitment (e.g., Ainslie, 1992). This construct is grounded on the wellknown finding that subjects in an intertemporal choice situation often chose to cancel in advance the availability of a SS alternative if given the opportunity (e.g., Rachlin & Green, 1972) therefore "committing" to undertake the delay associated with the LL alternative. However, in most intertemporal choice situations there are no opportunities to entirely prevent the availability of a temptation (i.e., SS alternative) that undermines the optimization of reward. Soft-commitment could be regarded as a behavior that reduces the probability of selecting the SS alternative without completely impeding it (Bryan, Karlan, & Nelson, 2010).
According to Rachlin (1995), when an individual refuses to take a SS alternative he/she may be avoiding the negative emotion associated with losing the opportunity to access the larger reward. For this to happen, the acquired aversiveness (i.e., negative emotion) associated with losing the larger reward should override the rewarding value of the smaller reward. Such acquired aversive properties that prevent impulsive choices could be conceived as an analogue of situations in which a specific action actually changes programmed contingencies to constraint or hinder the availability of temptations. This approach is entirely compatible with the framework offered here.
Some evidence that seems to be consistent with the view that soft-commitment affects the development of self-control comes from studies of delay of gratification in children. Mischel and Mischel (1983) found that in this protocol children 5 years and older (unlike younger children) often choose to cover the reward at the waiting period, what presumably decreases its saliency and, therefore, defection rate. Moreover, other studies have shown that children who employ strategies such as looking away or distracting themselves from the reward source show greater tolerance to delay (e.g., Rodriguez, Mischel, & Shoda, 1989). Looking away from or covering the reward source in this situation could be conceived as escaping a stimulus that controls defecting responses. The aversive properties involved in this presumed escape behavior would be acquired due to all the factors that we mentioned previously (expectancy violation, negative contingency, and time-out signaling), which is similar to the findings of Wasserman et al. (1974). At the risk of being too speculative, a tentative conclusion would be that conditioned inhibition may be involved in the process referred to as soft-commitment.
The Study of Impulsivity/Self-Control with Aversive Outcomes
Most studies of impulsivity involve appetitive outcomes. However, classic definitions do not imply that the impulsivity/self-control dimension should be limited to behavior in appetitive situations. Rachlin (1974) asserted that the delay-discounting account can be extended to address impulsiveness in aversive situations. He suggested that avoidance of a future larger aversive outcome by accepting a smaller aversive outcome in the present could be regarded as an index of self-control. Conversely, avoiding the smaller aversive outcome in the present, thus receiving the larger aversive outcome in the future, would be regarded as an index of impulsivity.
Deluty (1978) devised a laboratory protocol to empirically reproduce the possibility put forward by Rachlin (1974). In that protocol, rats had the opportunity to respond and undertake a brief electric shock. However, if they did not respond during a specified time window the animals lost the opportunity to respond and were then given a longer electric shock after a delay interval. Thus, subjects could maximize benefits by enduring a mildly noxious event. It is interesting that Deluty (1978; see also Deluty, Whitehouse, Mellitz, & Hineline, 1983) observed commitment responses in these conditions; if given the opportunity, some animals canceled in advance the opportunity to passively avoid the immediate shock. Deluty's protocol entails one of the most important conditions to test impulsivity espoused by the delaydiscounting approach, that is to establish a "now" versus "later" dilemma (Rachlin, 1995). However, it has an important difference with respect to intertemporal choice procedures, which are the "gold standard" in measuring delay discounting. Deluty's procedure does not imply a symmetric choice in which both alternatives are chosen by responses that are topographically similar. Instead, this protocol imposes a choice situation in which the two options involve responding and nonresponding. This can be considered as a rather asymmetrical choice situation, like go/no-go, stop-signal, and delay-of-gratification protocols.
We can readily identify aspects known to induce conditioned inhibition in this kind of procedure. Given the conditions that constitute Deluty's protocol, subjects might experience an expectancy violation when experiencing a brief electric shock in a context in which they sometimes receive longer electric shocks. Subjects also might be sensitive to the negative contingency between the brief electric shock and the long electric shock, because they never cooccur. Finally, subjects might leam that the brief electric shock signals a relatively long period without experiencing the large shock. Assuming that an inhibitor of an aversive US carries appetitive properties (Savastano et al., 1999), stimuli related with the brief shock would increase the frequency of the response that produces them. However, there are no stimuli explicitly arranged to signal the small shock in this protocol. A reasonable possibility is that the response's feedback (i.e., stimuli inevitably produced by the self-controlled response in this case) could acquire those appetitive properties (Dinsmoor, 2001). Whenever the acquired appetitive properties of a response's feedback override the inherent aversive properties of the brief shock, the individual would display self-controlled behavior.
Although avoiding a small aversive outcome at the expense of bringing about a larger aversive outcome in the future might reflect impulsivity, such a tendency has been suggested to reflect instead another category of maladaptive behavior. For example, Williams, Johns, and Norton (1998) attempted to explain anxiety-related psychiatric disorders as a deficit in conditioned-inhibition learning, which would also predict that anxious individuals would perform poorly in Deluty's protocol. One important difference between such an approach and ours is the motivational system involved in the hypothesized learning deficit. Williams et al. (1998) characterize anxiety as dysfunction in learning negative contingencies between a cue and an aversive stimulus. We propose that problems in learning that a cue is a negative predictor of a reward are crucial for the manifestation of impulsive behavior. In fact, there is considerable comorbidity between ADHD (one of which major symptoms is impulsivity) and anxiety disorders (see Schatz & Rostain, 2006). This would be compatible with the idea of a general conditioned-inhibition learning process, for which deficits can cause maladaptive impulsive-like behavior, anxiety-like behavior, or both. In fact, Nigg (2001) makes a similar proposition: he states that both ADHD and anxiety arise from a disinhibition problem. Nonetheless, he makes a contrasting distinction between the types of disinhibition involved in each case: ADHD would reflect an executive disinhibition problem, whereas anxiety would reflect a motivational disinhibition problem. More research on this topic would determine which of these taxonomies is more accurate.
Available Supporting Evidence and Further Tests for the Relationship between Conditioned Inhibition and Impulsivity
Appealing to conditioned inhibition to characterize some aspects of impulsive behavior may be fruitful, because there is a considerable amount of theoretical work devoted to conditioned inhibition (see Williams et al., 1992). This approach might inspire new research to test if the factors that influence conditioned inhibition have similar effects on impulsive/self-controlled behavior or vice versa. For example, one could predict that conditioned inhibition might be correlated with the metabolic rate of a species, as has been reported for impulsive choice (Tobin & Logue, 1994). Another prediction may be that learned self-control (i.e., the opposite tendency of impulsivity) would be a context-dependent phenomenon if it is learned second, as with conditioned inhibition (e.g., Miguez, Soares, & Miller, 2015; Nelson, 2002). It is interesting that there are already sources of evidence that support the hypothesis that conditioned inhibition and other measures of impulsivity are similarly affected by some biological factors. For example, both conditioned inhibition (Meyer & Bucci, 2014b) and preference for delayed rewards (Simon et al., 2010) are related to advanced stages in the ontogenetic development of rats. Another illustrative example is that the administration of d-amphetamine has a similar attenuating effect on impulsive behavior as measured by intertemporal choice, go/ no-go, SSTs (in humans; de Wit, Enggasser, & Richards, 2002), and conditioned-inhibition protocols (in rats; Harmer & Phillips, 1999). There is also evidence that the ablation of structures in the prefrontal cortex reduce both conditioned inhibition (medial prefrontal cortex; Meyer & Bucci, 2014a) and preference for delayed rewards (orbital prefrontal cortex; Mobini et al., 2002; but see Zeeb, Floresco, & Winstanley, 2010). These findings provide conditioned inhibition with face validity as a measure of impulsivity. However, all of these sources of evidence could be considered as indirect evidence, as they do not directly relate conditioned inhibition and the other measures of impulsivity.
Direct evidence for the relation of conditioned inhibition and other paradigms of impulsivity would be obtained from (1) studies that evaluate the correlation between conditioned inhibition and the other measures, (2) studies showing that training in one paradigm transfers to another paradigm, and (3) studies showing that there is a conditioned-inhibition function in stimuli embedded in the other paradigms. As far as we know, there are not yet published data from any of the first two types of studies. Regarding the latter kind of studies, as we have mentioned previously, Laude et al.'s (2014) and Trujano et al.'s (2016) studies suggested that a conditioned-inhibition process is (at least partially) responsible for optimal responding in a probabilistic choice paradigm, which has been considered as a paradigm to evaluate risk proneness-associated impulsivity.
It should be noted that the above-mentioned sources of evidence (i.e., Laude et al., 2014; Trujano et al., 2016), as they have been propounded, are restricted to assess the possibility that contingencies arranged under impulsivity tasks might give rise to conditioned inhibition. If this happens to be the case, another important consideration is whether conditioned inhibition plays a critical role in performance under tasks to measure impulsivity or its role is just incidental. A research agenda should include tactics to test whether or not the role of conditioned inhibition is crucial for the expression of impulsivity and/or for the development of self-control. For example, analytical methods designed to disentangle experimental effects from individual differences (e.g., Olejnik & Algina, 2003) would be suitable to consider the influence of withinindividual, between-individual, and between-group sources of variation (and interactions between those factors) in the behavioral indexes under study.
As we mentioned previously, some studies have reported poor or null correlations between performances on different categories of procedures to measure of impulsivity (see Evenden, 1999; Sonuga-Barke, 2005; Reynolds et al., 2006; Winstanley et al., 2004, 2006), which appears to undermine their predictive validity if one consider impulsivity as a monolithic construct. This may lead some authors to conclude that the processes that govern performance are not shared by the different procedures. However, there are even cases in which a number of measures within the same subcategory of procedures do not correlate (e.g., Bracks, Marshall-Pescini, Wallis, Huber, & Range, 2017). It makes no sense that measures of impulsivity are more prone to correlate with measures outside its domain than to correlate with measures within its subdomain. It seems hard to claim that there is an independent process that underlies every single behavioral manifestation of impulsivity.
Indeed, there are some alternative explanations for this common outcome. It may be possible that the tasks under study do not produce reliable interindividual variability, which is needed so that a sufficient statistical power can be achieved for detecting the association between two variables. According to Hedge, Powell, and Summer (2017), this is often the case for studies of individual differences studying relations between different behavioral tasks. The main problem is that the tasks used in these studies to measure a behavioral index are frequently devised in experimental settings, which paradoxically aim to minimize interindividual variability in order to increase statistical power. This issue might explain the large number of poor correlations reported when the relationship between two tasks of impulsivity is inquired, given that many such tasks have been adopted from an experimental tradition. Nakagawa and Schielzeth (2010) advocate that results of correlational studies should include estimates of interindividual variability (e.g., intra-class correlations) in order to construct a meaningful empirical synthesis of research in a given topic. This would be an appropriate practice for impulsivity research so as to inform the conclusions about comparisons between studies. Only if studies with an adequate statistical power confirm the absence of correlations between tasks, then one could claim the absence of a shared process.
However, yet another possible explanation for poor or null correlations between measures of impulsivity is that the putative shared process contributes relatively more to the state than to the trait facet of this behavioral dimension (see Odum & Bauman, 2010). If so, between-individuals correlations would not detect the association between two tasks. Longitudinal within-individuals correlations may complement the more common transversal between-individuals correlations (Kievit, Frankenhuis, Waldorp, & Borsboom, 2013) in order to draw inferences about the contribution of conditioned inhibition to state and not just the trait facets of impulsivity. Unless we test this possibility, a shared process between different forms of impulsivity cannot be ruled out altogether.
A number of authors have considered that inhibition deficit is an important part of impulsive behavior, ranging from those proposing that such a deficit is a fundamental feature of all manifestations of impulsivity (e.g., Bari & Robins, 2013; Buss & Plomin, 1975) to those that consider it to be just one of the various components of impulsive behavior (e.g., Richards et al., 2011). In this article, we provide a rationale that considers that factors influencing conditioned inhibition are present in tasks not necessarily designed to measure inhibitory processes (e.g., intertemporal choice), implying that these factors may have some impact on the subjects' performance in these tasks. If these assertions are correct, then studying conditioned inhibition as a learning process and its deficit as a behavioral trait would be of great importance for understanding of impulsive behavior. Some (e.g., Donahoe & Palmer, 1988; Papini & Bitterman, 1993; Skinner, 1938) have criticized the concept of inhibition as a behavioral phenomenon, which presumably has discouraged the study of this construct in some fields of behavior science. Here we reviewed some evidence and arguments that suggest that the behavioral phenomena under the rubric of conditioned inhibition possess enough validity to be worth studying. We argued that other plausible causes of the relative disregard of conditioned inhibition as a measure of impulsivity are the perceived necessity of applying summation and retardation test in every study of conditioned inhibition, and the apparent distance between this phenomenon and the prevailing definitions of response inhibition. We provided some examples that illustrate that the two-test issue could be circumvented. Regarding the definition issue, we highlighted some details in behavior controlled by conditioned inhibitors that could support the view that the distinctive features of other response inhibition tasks are nuances rather than fundamental differences.
Our approach may be related to efforts to characterize different impulsivity paradigms within the same framework. For example, Monterosso and Ainslie (1999) proposed that the underlying feature of impulsive behavior in several paradigms is a steep delay discounting. In a similar vein, Logue (1988) asserted that behavior in most of the paradigms designed to study impulsivity can be regarded as trade-off between smaller-sooner and larger-later rewards. On the other hand, Poulos et al. (1995) described choice impulsivity as a problem of disinhibition. One advantage of our approach is that we have identified, firmly based in learning theory, crucial aspects of conditioned inhibition that could be found in some paradigms of impulsivity, and specified how those aspects would influence performance in each case. This approach poses a significant challenge for learning theorists wishing to test its assumptions empirically; well-crafted research designs and data analysis methods would be required to test this hypothesis.
Acknowledgements The preparation of this manuscript was supported by the Direccion General de Asuntos del Personal Academico (DGAPA) postdoctoral fellowship grant (UNAM, 2016-2017) and by the Sistema Nacional de Investigadores (SNI) grant (file 64324). Special thanks to the Facultad de Psicologia and the Laboratorio de Mecanismos Neurales y Cognitivos del Aprendizaje. We thank Daniel Pearce, Nicole Muszynski, and Fred Westbrook for their helpful comments on a previous version of the manuscript.
Compliance with Ethical Standards
Declaration of Conflict of Interest Rodrigo Sosa and Cristiano V. dos Santos declare that they have no conflict of interest; we have no financial relationship with the organization that sponsored the research.
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Acebes, F., Solar, P., Moris, J., & Loy, I. (2012). Associative learning phenomena in the snail (Helix aspersa): Conditioned inhibition. Learning & Behavior, 40(1), 34-41.
Ainslie, G. (1992). Picoeconomics: The interaction of successive motivational states within the person. Cambridge: Cambridge University Press.
Alderson, R. M., Rapport, M. D., Sarver, D. E., & Kofler, M. J. (2008). ADHD and behavioral inhibition: A re-examination of the stopsignal task. Journal of Abnormal Child Psychology, 36, 989-998.
Alonso, E., Fairbank, M., & Mondragon, E. (2012). Conditioning for least action. In N. RuBwinkel, U. Drewitz, & H. van Rijn (Eds.), Proceedings 11th International Conference on Cognitive Modeling (ICCM-12), pp. 234-239. Berlin: Universitatsverlag der TU Berlin.
American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: American Psychiatric Publishing.
Anker, J. J., Zlebnik, N. E., Gliddon, L. A., & Carroll, M. E. (2009). Performance under a go/no-go task in rats selected for high and low impulsivity with a delay-discounting procedure. Behavioural Pharmacology, 20, 406-414.
Anselme, P. (2015). Incentive salience attribution under reward uncertainty: A Pavlovian model. Behavioural Processes, 111, 6-18.
Bardo, M. T., Cain, M. E., & Bylica, K. E. (2006). Effect of amphetamine on response inhibition in rats showing high or low response to novelty. Pharmacology Biochemistry & Behavior, 55(1), 98-104.
Bari, A., & Robins, T. W. (2013). Inhibition and impulsivity: Behavioral and neural basis of response control. Progress in Neurobiology, 108, 44-79.
Batson, J. D., & Best, M. R. (1981). Single-element assessment of conditioned inhibition. Bulletin of the Psychonomic Society, 18(6), 328330.
Bouton, M. E., Trask, S., & Carranza-Jasso, R. (2016). Learning to inhibit the response during instrumental (operant) extinction. Journal of Experimental Psychology: Animal Learning & Cognition, 42(3), 246-258.
Bracks, D., Marshall-Pescini, S., Wallis, L. J., Huber, L., & Range, F. (2017). Measures of dogs' inhibitory control abilities do not correlate across tasks. Frontiers in Psychology, 8, 849.
Brunton, C. II. (1883). On the nature of inhibition, and the action of drugs upon it. Nature, 27, 419-422.
Bryan, G., Karlan, D., & Nelson, S. (2010). Commitment devices. Annual Review of Economics, 2(1), 671-698.
Bucci, D. J., Hopkins, M. E., Keene, C. S., Sharma, M., & Orr, L. E. (2008). Sex differences in learning and inhibition in spontaneously hypertensive rats. Behavior & Brain Research, 187(1), 27-32.
Buss, A. H., & Plomin, R. (1975). A temperament theory of personality development. New York: Wiley.
Christianson, J. P., Benison, A. M., Jennings, J., Sandsmark, E. K., Amat, J., Kaufman, R. D. ... Maier, S. F. (2008). The sensory insular cortex mediates the stress-buffering effects of safety signals but not behavioral control. Journal of Neuroscience, 28(50), 1370313711.
Cole, R. P., Bamet, R. C., & Miller, R. R. (1997). An evaluation of conditioned inhibition as defined by Rescorla's two test strategy. Learning & Motivation, 28, 323-341.
de Wit, H., Enggasser, J. L., & Richards, J. B. (2002). Acute administration of d-amphetamine decreases impulsivity in healthy volunteers. Neumpsychopharmacology, 27(5), 813-825.
Deluty, M. Z. (1978). Self-control and impulsiveness involving aversive events. Journal of Experimental Psychology: Animal Behavior Processes, 4, 250-266.
Deluty, M. Z., Whitehouse, W. G., Mellitz, M., & Hineline, P. N. (1983). Self-control and commitment involving aversive events. Behaviour Analysis Letters, 3, 213-219.
Dinsmoor, J. A. (2001). Stimuli inevitably generated by behavior that avoids electric shock are inherently reinforcing. Journal of the Experimental Analysis of Behavior, 75(3), 311-333.
Donahoe, J. W., & Burgos, J. E. (1999). Timing without a timer. Journal of the Experimental Analysis of Behavior, 71(2), 257-263.
Donahoe, J. W., & Palmer, D. C. (1988). Inhibition: A cautionary tale. Journal of the Experimental Analysis of Behavior, 50(2), 333-341.
Dougherty, D. M., Bjork, J. M., Harper, R. A., Marsh, D. M., Moeller, F. G., Mathias, C. W., & Swann, A. C. (2003). Behavioral impulsivity paradigms: A comparison in hospitalized adolescents with disruptive behavior disorders. Journal of Child Psychology & Psychiatry, 44(8), 1145-1157.
Droungas, A., & LoLordo, V. M. (1994). Evidence of simultaneous excitatory and inhibitory associations in the explicitly unpaired procedure. Learning & Motivation, 25(1), 1-25.
El Massioui, N., Lamirault, C., Yagiie, S., Adjeroud, N., Garces, D., Maillard, A., ... Doyere, V. (2016). Impaired decision making and loss of inhibitory-control in a rat model of Huntington disease. Frontiers in Behavioral Neuroscience, 10, 204. https://doi.org/10. 3389/fiibeh.2016.00204.
Enticott, P. G., Ogloff. J. R. P., & Bradshaw, J. L. (2008). Response inhibition and impulsivity in schizophrenia. Psychiatry Research, 157, 251-254.
Evenden, J. L. (1999). Varieties of impulsivity. Psychopharmacology, 146, 348-361.
Fernando, A. B. P., Urcelay, G. P., Mar, A. C., Dickinson, A., & Robbins, T. W. (2014). Safety signals as instrumental reinforcers during freeoperant avoidance. Learning & Memory, 21, 488-497.
Gallistel, C. R., & Gibbon, J. (2000). Time, rate, and conditioning. The Psychological Review, 107(2), 289-344.
Gershon, J. (2002). A meta-analytic review of gender differences in ADHD. Journal of Attention Disorders, 5, 143-154.
Green, J. T., Chess, A. C., Conquest, C. J., & Yegla, B. A. (2011). Conditioned inhibition in a rodent model of attention-deficit/hyperactivity disorder. Behavioral Neuroscience, 125(6), 979-987.
Green. L., & Myerson, J. (2010). Experimental and correlational analyses of delay and probability discounting. In G. J. Madden & W. K. Bickel (Eds.), Impulsivity: The behavioral and neurological science of discounting (pp. 67-92). Washington, DC: American Psychological Association.
Green, L., & Myerson, J. (2013). How many impulsivities? A discounting perspective. Journal of the Experimental Analysis of Behavior, 99(1), 3-13.
Haddad, A. D., Pritchett, D., Lissek, S., & Lau, J. Y. (2012). Trait anxiety and fear responses to safety cues: Stimulus generalization or sensitization? Journal of Psvchopathology and Behavioral Assessment, 34(2), 323-331.
Haddon, J. E., & Killcross, A. S. (2005). Medial prefrontal cortex lesions abolish contextual control of competing responses. Journal of the Experimental Analysis of Behavior, 84, 485-504.
Harmer, C. J., & Phillips, G. D. (1999). Enhanced conditioned inhibition following repeated pretreatment with d-amphetamine. Psychopharmacology, 142, 120-131.
Harris, J. A., Kwok, D. W. S., & Andrew, B. J. (2014). Conditioned inhibition and reinforcement rate. Journal of Experimental Psychology: Animal Behavior & Cognition, 40(3), 335-354.
He, Z., Cassaday, H. J., Bonardi, C., & Bibby, P. A. (2013). Do personality traits predict individual differences in excitatory and inhibitory learning? Frontiers in Psychology, 4, 1-12.
He, Z., Cassaday, H. J., Howard, R. C., Khalifa, N., & Bonardi, C. M. (2011). Impaired Pavlovian conditioned inhibition in offenders with personality disorders. Quarterly Journal of Experimental Psychology, 64, 2334-2351.
Hearst, E., Besley, S., & Farthing, G. W. (1970). Inhibition and the stimulus control of operant behavior. Journal of the Experimental Analysis of Behavior, 3(2), 373-409.
Hearst, E., & Franklin, S. R. (1977). Positive and negative relations between a signal and food: Approach-withdrawal behavior to the signal. Journal of Experimental Psychology: Animal Behavior Processes, 3, 37-52.
Hedge, C., Powell. G., & Summer, P. (2017). The reliability paradox: Why robust cognitive tasks do not produce reliable individual differences. Behavioral Research Methods, 1-21.
Holland, P. C. (1979). Differential effects of omission contingencies on various components of Pavlovian appetitive conditioned responding in rats. Journal of Experimental Psychology: Animal Behavior Processes, 5(2), 178.
Horn, N. R., Dolan, M., Elliott, R., Deakin, J. F. W., & Woodruff, P. W. R. (2003). Response inhibition and impulsivity: An fMRI study. Neuropsychologica, 41, 1959-1966.
Huddy, V. C., Clark, L., Harrison, I., Ron, M. A., Moutoussis, M., Barnes, T. R. E., & Joyce, E. M. (2013). Reflection impulsivity and response inhibition in first-episode psychosis: Relationship to cannabis use. Psychological Medicine, 43, 2097-2107.
Iaboni, F., Douglas, V., & Baker, A. G. (1995). Effects of reward and response costs on inhibition in ADHD children. Journal of Abnormal Psychology, 104(1), 232-240.
Izquierdo, A., Belcher, A. M., Scott, L., Cazares, V. A., Chen, J., O'Dell, S. J., ... Marshall, J. F. (2010). Reversal-specific learning impairments after a binge regimen of methamphetamine in rats: Possible involvement of striatal dopamine. Neuropsychopharmacology, 35, 505-514.
Kievit, R. A., Frankenhuis, W. E., Waldorp, L. J., & Borsboom, D. (2013). Simpson's paradox in psychological science: A practical guide. Frontiers in Psychology, 4, 1-14.
Killeen, P. R. (1978). Stability criteria. Journal of the Experimental Analysis of Behavior, 29(1), 17-25.
Killeen, P. R. (2015). Models of ADHD: Five ways smaller sooner is better. Journal of Neuroscience Methods, 252, 2-13.
Killeen, P. R., & Fetterman, J. G. (1988). A behavioral theory of timing. Psychological Review, 95(2), 274-295.
Kramer, T. J., & Rilling, M. (1970). Differential reinforcement of low rates: A selective critique. Psychological Bulletin, 74(A), 225-254.
Laude, J. R., Stagner, J. P., & Zentall, T. R. (2014). Suboptimal choice by pigeons may result from the diminishing effect of nonreinforcement. Journal of Experimental Psychology: Animal Learning & Cognition, 40, 12-21.
Lawrence, A., Luty, J., Bogdan, N. A., Sahakian, B. J., & Clark, L. (2009). Impulsivity and response inhibition in alcohol dependence and problem gambling. Psychopharmacology, 207, 163-172.
Leitenberg, H. (1965). Is time out from positive reinforcement an aversive event? Psychological Bulletin, 64(6), 428-441.
Liu, S., Heitz, R. P., & Bradberry, C. W. (2009). A touch screen based stop-signal response task in rhesus monkeys for studying impulsivity associated with chronic cocaine self-administration. Journal of Neuroscience Methods, 777(1), 67-72.
Logan. G. D., Schachar, R. J., & Tannock, R. (1997). Impulsivity and inhibitory control. Psychological Science, 8, 60-64.
Logue, A. W. (1988). Research on self-control: An integrating framework. Behavior & Brain Science, II, 665-709.
Logue, A. W. (1995). Self-control: Waiting until tomorrow for what you want today. Englewood Cliffs: Prentice Hall.
Lohr, J. M., Olatunji, B. 0., & Sawchuk, C. N. (2007). A functional analysis of danger and safety signals in anxiety disorders. Clinical Psychology Review, 27, 114-126.
LoLordo, V. M. (1969). Positive conditioned reinforcement from aversive situations. Psychological Bulletin, 72, 193-203.
LoLordo, V. M., & Fairless, J. L. (1985). Pavlovian conditioned inhibition: The literature since 1969. In R. R. Miller & N. E. Spear (Eds.), Information processing in animals: Conditioned inhibition (pp. 149). Hillsdale: Lawrence Erlbaum Associates.
Lopez, P., Alba. R., & Orduna, V. (2017). Individual differences in incentive salience attribution are not related to suboptimal choice in rats. Behavioural Brain Research, 341(2), 71-78.
MacKillop, J., Weafer, J., Gray, J. C., Oshri, A., Palmer, A., & de Wit. H. (2016). The latent structure on impulsivity: Impulsive choice, impulsive action, and impulsive personality traits. Psychopharmacology, 235(18), 3361-3370.
Mackintosh, N. J., & Cotton, M. M. (1985). Conditioned inhibition from reinforcement reduction. In R. R. Miller & N. E. Spear (Eds.), Information processing in animals: Conditioned inhibition (pp. 89-111). Hillsdale: Lawrence Erlbaum Associates.
MacLeod, C. M., Dodd, M. D., Sheard, E. D., Wilson, D. E., & Bibi, U. (2003). In opposition to inhibition. Psychology of Learning & Motivation, 43, 163-215.
MacLeod, J. E., Potter, A. S., Simoni, M. K, & Bucci, D. J. (2006). Nicotine administration enhances conditioned inhibition in rats. European Journal of Pharmacology, 551, 76-79.
Madden, G. J., & Bickel, W. K. (2010). Impulsivity: The behavioral and neurological science of discounting. Washington. DC: American Psychological Association.
Meyer, H. C., & Bucci, D. J. (2014a). The contribution of medial prefrontal cortical regions to conditioned inhibition. Behavioral Neuroscience, 128(6), 644-653.
Meyer, H. C., & Bucci, D. J. (2014b). The ontogeny of learned inhibition. Learning & Memory, 21, 143-152.
Miguez, G., Soares, J. S., & Miller, R. (2015). The role of test context in latent inhibition of conditioned inhibition: Part of a search for general principles of associative interference. Learning & Behavior, 43(3), 228-242.
Miller, R. R., Hallam, S. C., Hong, J. Y., & Dufore, D. S. (1991). Associative structure of differential inhibition: Implications for models of conditioned inhibition. Journal of Experimental Psychology: Animal Behavior Processes, 17(2), 141.
Miller, R. R., & Matzel, L. D. (1988). The comparator hypothesis: A response rule for the expression of associations. The Psychology of Learning & Motivation, 22, 51-92.
Mischel, H. N., & Mischel, W. (1983). The development of children's knowledge of self-control strategies. Child Development, 54, 603-619.
Mischel, W., Shoda, Y., & Rodriguez, M. (1989). Delay of gratification in children. Science, 244, 933-938.
Mobini, S., Body, S., Ho, M. Y., Bradshaw, C. M., Szabadi, E., Deakin, J. F. W., & Anderson, I. M. (2002). Effects of lesions of the orbitofrontal cortex on sensitivity to delayed and probabilistic reinforcement. Psychopharmacology, 160, 290-298.
Monterosso, J., & Ainslie, G. W. (1999). Beyond discounting: Possible experimental models of impulse control. Psychopharmacology, 146, 339-347.
Moscovitch, A., & LoLordo, V. M. (1968). Role of safety in the Pavlovian backward fear conditioning procedure. Journal of Comparative & Physiological Psychology, 66(3pl), 673.
Mowrer, O. H. (1956). Two-factor learning theory reconsidered, with special reference to secondary reinforcement and the concept of habit. Psychological Review, 63(2), 114-128.
Nakagawa, S., & Schielzeth, H. (2010). Repeatability for Gaussian and non-Gaussian data: A practical guide for biologists. Biological Reviews, 85, 935-956.
Navarra, R., Comery, T. A., Graf, R., Rosensweig-Lipson, S., & Day, M. (2007). The 5HTC2c receptor agonist WAY-163909 decreases impulsivity in the 5-choice serial reaction time test. Behavioural Brain Research, 188(2), 412-415.
Nelson, J. B. (2002). Context specificity of excitation and inhibition in ambiguous stimuli. Learning & Motivation, 33, 284-310.
Nigg, J. T. (2001). Is ADHD a disinhibitory disorder? Psychological Bulletin, 127(5), 571-598.
Nombela, C., Rittman. T., Robbins, T. W., & Rowe, J. B. (2014). Multiple modes of impulsivity in Parkinson's disease. PLoS One, 9(1), e85747.
Odum, A. L., & Bauman, A. A. L. (2010). Delay discounting: State and trait variable. In G. J. Madden & W. K. Bickel (Eds.), Impulsivity: The behavioral and neurological science of discounting (pp. 3966). Washington, DC: American Psychological Association.
Olejnik, S., & Algina, J. (2003). Generalized eta and omega squared statistics: Measures of effect size for some common research designs. Psychological Methods, 8(4), 434-447.
Orduna, V., Valencia-Torres, L., & Bouzas, A. (2009). DRL performance of spontaneously hypertensive rats: Dissociation of timing and inhibition of responses. Behavior Brain Research, 201, 158-165.
Papini. M. R. (1988). Associative learning in the marsupials Didelphis albiventris and Lutreolina crassicaudata. Journal of Comparative Psychology, 102, 21-27.
Papini, M. R. (2001). Comparative psychology: Evolution and development of behavior. Upper Saddle River: Prentice Hall.
Papini, M. R., & Bitterman, M. E. (1993). The two-test strategy in the study of inhibitory conditioning. Journal of Experimental Psychology: Animal Behavior Processes, 19(4), 342-352.
Papini, M. R., & Dudley, R. T. (1997). Consequences of surprising reward omissions. Review of General Psychology, 1(2), 175-197.
Papini, M. R., & White, N. (1994). Performance during signals for reward omission. Learning and Motivation, 25, 45-64.
Pavlov, I. P. (1927). Conditioned reflexes. Oxford: Oxford University Press.
Poulos, C. X., Parker, J. L., & Le, A. D. (1995). Impulsivity predicts individual susceptibility to high levels of alcohol self-administration. Behavioural Pharmacology, 6, 810-814.
Rachlin, H. (1974). Self-control. Behaviorism, 2(1), 94-107.
Rachlin, H. (1995). Self-control: Beyond commitment. Behavioral & Brain Sciences, 18, 109-159.
Rachlin, H., & Green, L. (1972). Commitment, choice, and self-control. Journal of the Experimental Analysis of Behavior, 17, 15-22.
Rachlin, H.. Kagel, J. H., & Battalio, R. C. (1980). Substitutability in time allocation. Psychological Review, 57, 355-374.
Rachlin, H., Rainieri, A., & Cross, D. (1991). Subjective probability and delay. Journal of the Experimental Analysis of Behavior, 55, 233-244.
Rescorla, R. A. (1969a). Conditioned inhibition of fear resulting from negative CS-US contingencies. Journal of Comparative & Physiological Psychology, 67(4), 504-509.
Rescorla, R. A. (1969b). Pavlovian conditioned inhibition. Psychological Bulletin, 72(2), 77-94.
Reynolds, B.. de Wit, H., & Richards, J. B. (2002). Delay of gratification and delay discounting in rats. Behavioural Processes, 59, 157-168.
Reynolds, B., Ortengren, A., Richards, J. B., & De Wit, H. (2006). Dimensions of impulsive behavior. Personality and behavioral measures. Personality & Individual Differences. 40, 305-315.
Reynolds, B., & Schiffbauer, R. (2005). Delay of gratification and delay discounting: A unifying feedback model of delay-related impulsive behavior. The Psychological Record, 55, 439-460.
Richards, J. B., Gancarz, A. M., & Hawk, L. W. (2011). Animal models of behavioral processes that underlie the occurrence of impulsive behaviors in humans. In M. T. Bardo, D. Fishbein, & R. Milich (Eds.), Inhibitory control and drug abuse prevention: From research to translation (pp. 13--41 >. New York, NY: Springer.
Richards, J. B., Sabol, K. E., & Seiden, L. S. (1993). DRL interresponsetime distributions: Quantification by peak deviation analysis. Journal of the Experimental Analysis of Behavior, 60, 361-385.
Robinson, E. S. J., Eagle, D. M., Economidou, D., Theobald, D. E. H., Mar, A. C., Murphy, E. R., ... Dalley, J. W. (2009) Behavioral characterization of high impulsivity on the 5-choice serial reaction time task: Specific deficits in "waiting" versus "stopping." Behavioural Brain Research, 196(2), 310-316.
Rodriguez, M. L., Mischel, W., & Shoda, Y. (1989). Cognitive person variables in the delay of gratification of older children at risk. Journal of Personality & Social Psychology, 57(2), 358-367.
Rodriguez, W., Bouzas, A., & Orduna, V. (2017). Temporal discounting of aversive consequences in rats. Learning & Behavior, 1-11.
Rosval, L., Steiger, H., Bruce, K., Israel, M., Richardson, J., & Aubut, M. (2006). Impulsivity in women with eating disorders: Problem of response inhibition, planning, or attention? International Journal of Eating Disorders, 39(1), 590-593.
Rushen, J. (1986). The validity of behavioural measures of aversion: A review. Applied Animal Behavior Science, 16, 309-323.
Russell, V. A., de Villiers, A., Sagvolden, T., Lamm, M., & Taljaard, J. (1995). Altered dopaminergic function in the prefrontal cortex, nucleus accumbens and caudate-putamen of an animal model of attention-deficit hyperactivity disorder: The spontaneously hypertensive rat. Brain Research, 676, 343-351.
Sanabria, F., & Killeen, P. R. (2008). Evidence for impulsivity in the spontaneously hypertensive rat drawn from a complementary response-withholding tasks. Behavioral & Brain Functions, 4, 724.
Savastano, H. I., Cole, R. P., Barnett, R. C., & Miller, R. R. (1999). Reconsidering conditioned inhibition. Learning & Motivation, 30, 101-127.
Schatz, D. B., & Rostain. A. L. (2006). ADHD with comorbid anxiety: A review of the current literature. Journal of Attention Disorders, 10, 141-149.
Scheres, A.. Oosterlaan, J., Swanson, J., Morein-Zamir, S., Meiran, N., Schut, H., ... Sergeant, J. A. (2003). The effect of methylphenidate on three forms of response inhibition in boys with AD/HD. Journal of Abnormal Child Psychology. 31(1), 105-120.
Simon, N. W., LaSarge, C. L., Montgomery, K. S., Williams, M. T., Mendez, I. A., Setlow, B., & Bizon, J. (2010). Good things come to those who wait: Attenuated discounting of delayed rewards in aged Fischer 344 rats. Neurobiology of Aging, 31, 853-862.
Skinner, B. F. (1938). The behavior of organisms. New York: AppletonCentury-Crofts.
Solanto, M. V., Abikoff, H., Sonuga-Barke, E., Schachar, R., Logan, G. D., Wigal, T.,... Turkel, E. (2001). The ecological validity of delay aversion and response inhibition of impulsivity in AD/HD: A supplement to the NIMH multimodal treatment study of AD/HD. Journal of Abnormal Child Psychology, 29(3), 215-228.
Sonuga-Barke, E. J. S. (2005). Causal models of attention-deficit/hyperactivity disorder: From common simple deficits to multiple developmental pathways. Biological Psychiatry, 57(11), 1231-1238.
Sosa, R., & dos Santos, C. V. (2014a). ?Pueden estimulos relacionados con la ausencia de recompensa reducir la impulsividad en ratas? Acta Comportamentalia, 22(2), 153-167.
Sosa, R., & dos Santos, C. V. (2014b). El papel de los reforzadores condicionados en la conducta autocontrolada. En M. Serrano (Ed.) La investigacion del comportamiento animal en Mexico: Teorias y estudios contemporaneos (pp. 99-113). Jalapa: Universidad Veracruzana.
Sosa, R., & dos Santos, C. V. (2018). Toward a unifying account of impulsivity and the development of self-control. Perspectives on Behavioral Science, 1(1), 1-32.
Stahl, C., Voss, A., Schmitz, F., Nuzbaum, M., Tuscher, O., Lieb, K., & Klauer, K. C. (2014). Behavioral components of impulsivity. Journal of Experimental Psychology: General, 143(2), 850-886.
Stewart, P. W. Sargent D. M., Reihman, J., Gump, B. B., Lonky, E., Darvill, T., ... Pagano, J. (2006). Response inhibition during differential reinforcement of low rates (DRL) schedules may be sensitive to low-level polychlorinated biphenil, methilmercury, and lead exposure in children. Children's Health: Environmental Health Perspectives, 114(12), 1923-1929.
Stout, S., Escobar, M., & Miller, R. R. (2004). Trial number and compound stimuli temporal relationship as joint determinants of second order conditioning and conditioned inhibition. Learning & Behavior, 32(2), 230-239.
Swann, A. C., Lijffijt, M., Lane, S. D., Steinberg, J. L., & Moeller, F. G. (2009). Trait impulsivity and response inhibition in antisocial personality disorder. Journal of Psychiatric Research, 43, 1057-1063.
Terrace, H. S. (1966). Discrimination learning and inhibition. Science, 154(3151), 1677-1680.
Tobin, H., & Logue, W. (1994). Self-control across species (Columba livia, Homo sapiens, and Rattus norvegicus). Journal of Comparative Psychology, 108(2), 126-133.
Tobler, P. N., Dickinson, A., & Schulz, W. (2003). Coding of predicted reward omission by dopamine neurons in a conditioned inhibition paradigm. Journal of Neuroscience, 23(32), 10402-10410.
Tomie, A., Aguado, A. S., Pohorecky, L. A., & Benjamin, D. (1998). Ethanol induces impulsive-like responding in a delay-of-reward operant choice procedure: Impulsivity predicts autoshaping. Psychopharmacology, 139, 376-382.
Trommer, B. L., Hoeppner, J. B., Lorber, R., & Armstrong, K. J. (1988). The go/no-go paradigm in attention deficit disorder. Annals of Neurology, 24(5), 610-614.
Trujano, R. E., Lopez, P.. Rojas-Leguizamon, M., & Ordufia, V. (2016). Optimal behavior by rats in a choice task is associated to a persistent conditioned inhibition effect. Behavioural Processes, 130, 65-70.
van den Broek, M. D., Bradshaw, C. M., & Szabadi, E. (1987). Behaviour of "impulsive" and "non-impulsive" humans in a temporal differentiation schedule of reinforcement. Personality & Individual Differences, 8(2), 233-239.
Vigil-Colet, A., & Codorniu-Raga, M. J. (2004). Aggression and inhibition deficits, the role of functional and dysfunctional impulsivity. Personality & Individual Differences, 37, 1431-1440.
Wagner, A. R., Mazur, J. E., Donegan, N. H., & Pfautz, P. L. (1980). Evaluation of blocking and conditioned inhibition to a CS signaling a decrease in US intensity. Journal of Experimental Psychology: Animal Behavior Processes, (5(4), 376-385.
Wagner, A. R., & Rescorla, R. A. (1972). Inhibition in Pavlovian conditioning: Application of a theory. In R. A. Boakes & M. S. Halliday (Eds.), Inhibition and learning (pp. 301-336). London: Academic Press.
Wasserman, E. A., Franklin, S., & Hearst, E. (1974). Pavlovian appetitive contingencies and approach vs. withdrawal to conditioned stimuli in pigeons. Journal of Comparative & Physiological Psychology, 86, 616-627.
Watterson, E., Mazur, G. J., & Sanabria, F. (2015). Validation of a method to assess ADHD-related impulsivity in animal models. Journal of Neuroscience Methods, 252, 36-47.
Wilbertz, T., Lorenz, D., Horstmann, A., Neumann, J., Villringer, A., Heinze, H., ... Schlagenhauf, F. (2014). Response inhibition and its relation to multidimensional impulsivity. Neuroimage, 103, 241-248.
Williams, D. A. (1996). A comparative analysis of negative contingency learning in humans and nonhumans. In D. R. Shanks, K. J. Holyoak, & D. L. Medin (Eds.), Causal learning (pp. 89-131). San Diego: Academic Press.
Williams, D. A., Johns, K. W., & Norton, G. R. (1998). Conditioned inhibition and its applications in panic and obsessive-compulsive disorders. In W. O'Donohue (Ed.), Learning & Behavior Therapy (pp. 85-106). Needham Heights: Allyn & Bacon.
Williams, D. A., Overmier, J. B., & Lolordo, V. M. (1992). A reevaluation of Rescorla's early dictums about conditioned inhibition. Psychological Bulletin, 111(2), 275-290.
Winstanley, C. A. (2011). The utility of rat models of impulsivity in developing pharmacotherapies for impulse control disorders. British Journal of Pharmacology, 164, 1301-1321.
Winstanley, C. A., Chudasama, Y., Dalley, J. W., Theobald, D. E. H., Glennon, J. C., & Robbins, T. W. (2003). Intra-prefrontal 8-OHDPAT and M100907 improve visuospatial attention and decrease impulsivity on the five-choice serial reaction task in rats. Psychopharmacology, 7(57,304-314.
Winstanley, C. A., Dalley, J. K., Theobald, D. E. H., & Robbins, T. W. (2004). Fractionating impulsivity: Contrasting effects of impulsive behavior. Neuropsychopharmacology, 29, 1331-1343.
Winstanley, C. A.. Eagle, D. M., & Robbins, T. W. (2006). Behavioral models of impulsivity in relation to ADHD: Translation between clinical and preclinical studies. Clinical Psychology Review, 26(4), 379-395.
Zeeb, F. D., Floresco, S. B., & Winstanley, C. J. (2010). Contributions of the orbitofrontal cortex to impulsive choice: Interactions with basal levels of impulsivity, dopamine signalling, and reward-related cues. Psychopharmacology, 27/(1), 87-98.
Zilio, D. (2016). On the autonomy of psychology from neuroscience: A case study of Skinner's radical behaviorism and behavior analysis. Review of General Psychology, 20(2), 155.
Rodrigo Sosa (1,2) [iD] * Cristiano Valerio dos Santos (3)
[mail] Rodrigo Sosa
Cristiano Valerio dos Santos
Published online: 15 November 2018
(1) Laboratorio de Neurociencias, Universidad Iberoamericana, Prolongation Paseo de la Reforma 880, Lomas de Santa Fe, Zip Code, 01219 Mexico City, Mexico
(2) Facultad de Psicologia, Universidad Nacional Autonoma de Mexico, Av. Universidad 3004, Copilco-Universidad, Zip Code, 04510 Mexico City, Mexico
(3) Centra de Estudios e Investigaciones en Comportamiento, Universidad de Guadalajara, Francisco de Quevedo 180, Arcos Vallarta, Zip Code, 44130 Guadalajara, Jalisco, Mexico
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||THEORETICAL ARTICLE|
|Author:||Sosa, Rodrigo; dos Santos, Cristiano Valerio|
|Publication:||The Psychological Record|
|Date:||Jun 1, 2019|
|Previous Article:||The On-Going Search for Perspective-Taking IRAPs: Exploring the Potential of the Natural Language-IRAP.|
|Next Article:||Identifying Accurate and Inaccurate Stimulus Relations: Human and Computer Learning.|