Rmac_vol_32-1.indd

REVISTA MEXICANA DE ANÁLISIS DE LA CONDUCTA 32, 73-91
SPECIAL INVITED ARTICLE/ARTÍCULO ESPECIAL INVITADO “BEHAVIOR-RELEASING” EFFECTS OF DRUGS:
ANTI-PUNISHMENT AND ANTI-CONFLICT
PROCEDURES
EFECTOS LIBERADORES CONDUCTUALES DE LAS DROGAS: PROCEDIMIENTOS DE ANTI-CASTIGO Y ANTI-CONFLICTO ERIN B. RASMUSSEN1
ABSTRACT
Anxiolytic drugs are said to reduce internal states (e.g., anxiety) that are indu-ced by the presentation of aversive events, leading to a release of behavior that has been suppressed. The objective of the present paper is to discuss two categories of procedures that are often used interchangeably-- anti-pu-nishment (procedures that release punished behavior) and anti-confl ict (pro-cedures that reduce avoidance behavior). Similarities and differences bet-ween the two categories of procedures are reviewed, emphasizing distinctio-ns between theoretical, methodological, and pharmacological specifi cs. The anxiety-as-explanation issue is discussed also, in terms of possibly obscuring the behavioral and pharmacological mechanisms that are involved with drugs that “release” behavior.
Key words: anxiolytic drugs, anti-punishment, anti-confl ict, anxiety, expla- nation, behavioral mechanisms, pharmacological mechanisms. 1. The author acknowledges with gratitude Chris Newland and Steven Lawyer for comments on earlier drafts. The author also thanks the editor, José E. Burgos, for his invitation to write this Se ha dicho que las drogas anxiolíticas reducen estados internos (e.g., an-siedad) que son inducidos por la presentación de eventos aversivos, y llevan a la liberación de conducta que ha sido suprimida. El objetivo del presente artículo es discutir estas dos categorías de procedimientos que con frecuen-cia se usan de manera intercambiable -- anti-castigo (procedimientos que liberan conducta castigada) y anti-confl icto (procedimientos que liberan con-ducta de evitación). Se revisan las semejanzas y diferencias entre las dos categorías, haciendo énfasis sobre detalles teóricos, metodológicos y farma-cológicos. También se discute el problema de la ansiedad como explicación, en términos del posible oscurecimiento que tal explicación produce sobre los mecanismos conductuales y farmacológicos involucrados en las drogas que “liberan” conducta.
Palabras clave: liberación de conducta, drogas, anti-castigo, anti-confl ic- to, ansiedad, explicación, mecanismos conductuales, mecanismos farmaco-lógicos. Anxiolytic (anxiety-reducing) drugs have been shown to selectively “release” behavior that has been suppressed. The basic procedure involves a choice between two alternatives. One alternative consists of a simultaneously appe-titive and aversive condition (also known as confl ict), for example, the concur-rent delivery of food and shock. The second alternative consists of avoidance of the fi rst alternative, i.e., engaging in “other” behavior. Under non-drug con-ditions, the human or animal tends to engage in the latter alternative. Under the anxiolytic drug, the former alternative is more infl uential on behavior, pre-sumably because behavior is less sensitive to the aversive condition. In the behavioral pharmacological literature, the terms anti-punishment and anti-confl ict are used to describe drug effects on behavior under approach-avoidance types of procedures. The terms are often used interchangeably (see, for example, Fontana, Carbary, & Commissaris, 1989; Koek & Cob-paert, 1991; Lerner, Feldon, & Myslobodsky, 1980; Pattij, Hijzen, Gommans, Maes, & Olivier, 2000; Thiebot, Soubrie, & Simon, 1985; Waddington & Olley, 1977), although on close inspection, important differences exist between the categories of the procedures. Anti-punishment effects refer to a drug-induced increase in behavior that specifi cally has been punished within the labora-tory setting. Typically, a baseline response rate of lever-pressing or other free operant is established under a schedule of food reinforcement. Electric cur-rent, or some other aversive stimulus, is then delivered under a schedule of punishment, reducing the rate of lever-pressing relative to the unpunished baseline. Alternative behavior, such as exploring or grooming, becomes more probabilistic. The anti-punishment effects of an anxiolytic drug are evident when an increase in behavior suppressed by punishment is observed, and this increase is independent of an increase in response rate in the unpun-ished baseline. In anti-confl ict procedures, an organism is presented with a situation that has both aversive and appetitive properties, though they are less clearly specifi ed than in anti-punishment procedures. Under non-drug conditions, the animal does not engage in the task, but when specifi c types of anxiolytic drugs are administered, the animal may engage in the task more frequently. For example, in the light-dark task a box is split into two compartments; one half is illuminated and one half is not. Under control conditions, rodents spend more time in the darkened area of the light-dark test box and are also less active in the illuminated portion of the box than in the darkened portion. Some anxiolytics appear to increase selectively the amount of time in illuminated areas. There is no experimental induction of punishment in the anti-confl ict situation, but anti-confl ict is similar to anti-punishment in that behavior that occurs at a low probability is released by the drug.
The tendency to confuse the terms anti-punishment and anti-confl ict pro- cedures likely comes from the common practice of categorizing and discuss-ing these procedures as animal (and human) models of anxiety. The con-ceptualization goes as follows: Engaging in a situation that is simultaneously appetitive and aversive induces anxiety or discomfort, and the administration of the drug is said to remove the subjective anxiety or discomfort. Hence, the mechanism for behavioral change is removal of anxiety or discomfort. In some instances, the mechanism refers to a disengaging of a behavioral inhi-bition system (BIS; Gray, 1988). The drawback is that conceptualizing anxiety or the BIS as the explanatory mechanism has attenuated the search for be-havioral mechanisms involved with anxiolytic drugs, which may include, for example, a drug-induced decrease in behavioral sensitivity to the aversive condition or perhaps a drug-induced increase in behavioral sensitivity to the appetitive condition. (Gray, 1988, specifi es three possible anxiolytic mecha-nisms: behavioral inhibition, preparation for vigorous action, and increased at-tention to the environment. However, the specifi c mechanisms of behavior for those three suggestions are not clearly defi ned.) MacCorquodale and Meehl (1948) warned that the use of labels, such as anxiety, can lead to construct reifi cation; that is, the term becomes explanatory (or used as a real cause) instead of descriptive, as it was originally intended. Moreover, the singular focus of anxiety as an explanatory variable has halted the discussion of be-havioral differences between anti-confl ict and anti-punishment procedures. This is interesting since some drugs have differential effects across proce-dures, suggesting different neurochemical substrates that might underlie two different behavioral processes, as opposed to two different types of anxiety as some researchers suggest (see, for example, Stefanski, Paleiko, Kostowski, & Plaznik, 1992). The purpose of this paper is to describe and summarize the literature on the releasing function that anxiolytic drugs have on behavior, and to elucidate the differences between anti-punishment and anti-confl ict proce-dures. Additionally, the place of anxiety in anti-confl ict and anti-punishment will be critiqued.
ANTI-PUNISHMENT
Anti-punishment effects are generally observed with drugs that are GABA agonists, such as barbiturates, benzodiazepines, and alcohol; opiates, stimu-lants, and other drugs do not generate anti-punishment effects. There is some inconsistent evidence that some serotonergic drugs (often used for the treat-ment of mood disorders) also have anti-punishment effects. These drugs ef-fects will be reviewed later. Anti-punishment effects have been demonstrated across a variety of spe- cies, including rats (e.g., Koob, Braestrup, & Britton, 1986; Vogel, 1980), squirrel monkeys (e.g., Barrett, Brady, & Witkin, 1985), cats (e.g., Masserman & Yum, 1946), pigeons (e.g., Brocco, Koek, Degruyse, & Colpaert; Koek & Coelpaert, 1991; Mansbach, Harrod, Hoffman, Nader, Lei, Witkin, & Barrett, 1988), and humans (e.g., Carlton, Siegel, Murphee, & Cook 1981; Rasmussen & New-land, 2006). These effects have been observed across a variety of procedures, as well. These procedures are summarized in Table 1 and discussed below. Conjoint Schedules of Reinforcement and Punishment and Conditioned Suppression Barrett, Brady, & Witkin (1985) placed lever-pressing of squirrel monkeys un-der a conjoint fi xed interval 3’ fi xed ratio 30-shock schedule, such that the fi rst response that occurred after every 3 min elapsed was reinforced with liquid reinforcement or a banana chip. In addition, every 30th response was punished with shock, and this contingency suppressed response rate. Doses of the GABAergic compounds chlordiazepoxide (1-60 mg/kg), pentobarbital (3-10 mg/kg) and ethanol (1-2.5 g/kg) increased punished response rates si-milarly (between 125 and 225% of punished control rates), but did not increa-se unpunished responding. Comparably strong anti-punishment effects have been replicated with chlordiazepoxide (Mansbach, Harrod, Hoffman, Nader, Lei, Witkin, & Barrett, 1988; Witkin, Mansbach, Barrett, Bolger, Skolnick, & Weissman, 1987) under similar schedules of reinforcement and punishment. Moderate anti-punishment effects have been observed under buspirone (Wit-kin, Mansbach, Barrett, Bolger, Skolnick, & Weissman, 1987), gepirone, clo- Table 1. Summary of studies demonstrating anti-punishment and anti- Strong Effects Shown with These Drugs chlordiazepox., diazepam, oxazepam, meprobamate, phenobarbital, secobarbital, pentobarbital, ethanol, alprazolam, alprazolam, perphanazine, chlordiazepoxide, propranol, phenobarbital, secobarbital, pentobarbital, ethanol, diazepam chlordiazepoxide, diazepam, meprobamate, phenobarbital, amobarbital, chlordiazepox., pentobarbital; ethanol; ipsapirone, PAPP, (-)MDL 72832, (+) MDL 72832, buspirone, Other (e.g., open fi eld activity, fl uoxetine, buspirone, mainserin, gepiron zapine, and 8-OH-DPAT (Mansbach, Harrod, Hoffman, Nader, Lei, Witkin, & Barrett, 1988), all of which are serotonergic agonists.
Dworkin, Bimle, & Miyauchi (1989) found results similar to the above- mentioned anti-punishment studies after placing bar-pressing under a ran-dom ratio (RR) schedule of reinforcement, in which a probability generator randomly determined whether the response would be reinforced after a num-ber of lever presses. Rats pressed an average of 55 responses per minute under this schedule. Then a RR schedule of shock was superimposed on the reinforcement schedules, and this suppressed bar-pressing by an order of magnitude. Low doses of pentobarbital (5.6 and 10 mg/kg) increased rates of punished responding by fi ve to 10 times, but did not increase unpunished responding. Cocaine, which normally increases low rates of responding main-tained by reinforcement, did not produce anti-punishment effects. Operant licking also has been used for examining anti-punishment ef- fects with conjoint schedules of reinforcement and punishment, and this is often referred to as the Vogel procedure (after Vogel, Beer, & Clody, 1971). Here, water-deprived rats make contact with a spout that produces water or some other liquid under a continuous reinforcement schedule. Once the lick-ing comes under the control of the liquid reinforcer, licking is punished by de-livery of electric current under a schedule of punishment. In the original study, Vogel, Beer, and Clody (1971), programmed a fi xed ratio 20 (FR20) schedule of shock (every twentieth lick produced a shock) and licking was suppressed to low levels. The GABAergic compounds chlordiazepoxide (6 and 8 mg/kg), diazepam (2 and 4 mg/kg), oxazepam (3 to 12.5 mg/kg), meprobamate (30 to 120 mg/kg) and pentobarbital (5 and 10 mg/kg) dose-dependently increased the rates of punished, but not unpunished, licking. d-Amphetamine (a cat-echolamine agonist), pemoline (an epileptic) and scopolamine (a cholinergic antagonist) produced no anti-punishment effects. Similar anti-punishment ef-fects with the Vogel task have been replicated with diazepam (e.g., Liljeq-uist and Engel, 1984; Stefanski, Palejko, Kostowski, & Plaznik, 1992) and chlordiazepoxide (e.g., Brocco, Koek, Degruyse, & Colpaert; Kennet, Trail, & Bright, 1998), and have been found with other GABAergic compounds, such as alprazolam (Kennet, Trail, & Bright, 1998), midazolam (Stefanski, Palejko, Kostowski, & Plaznik, 1992), and ethanol (Liljequist and Engel, 1984; Vogel, 1980). Buspirone yields confl icting data in which anti-punishment effects are inconsistently demonstrated, but when they are, they are moderate at best (cf. Kennet, Trail, & Bright, 1998; Schefke, Fontana, & Commissaris, 1989; Stefanski, Palejko, Kostowski, & Plaznik, 1992). Ritaserin, ipsarone, and vari-ous other serotonergic drugs were shown to increase punished responding to a moderate extent (Stefanski, Palejko, Kostowski, & Plaznik, 1992); however it is unclear whether this increase was specifi c to punished responding, be-cause data on unpunished responding were not shown. Hence, it is uncertain whether these were true anti-punishment effects, or are able to be explained by other behavioral mechanisms, such as rate dependence. Operant licking also has been used to examine anti-punishment effects within the context of the conditioned suppression procedure (see Estes & Skinner, 1941). In the conditioned suppression procedure, an aversive, usu-ally shock, is paired with a stimulus. The presentation of the conditioned aversive stimulus alone (without the shock) induces a suppression of be-havior, relative to the absence of the conditioned stimulus. Conditioned sup-pression of bar-pressing and conditioned suppression of drinking (CSD), in which the operant is licking a water spout, has been demonstrated with di-azepam (Commissaris, Harrington, & Altman, 1990; Kilts, Commissaris, Mc-Closkey, Damian, Brown, Barraco, & Altman, 1990; Commissaris & Rech, 1981; McCloskey, Paul, & Commissaris, 1987; Mokler & Rech, 1985), al- prazolam (Commissaris, Harrington, & Altman, 1990); phenobarbital (Com-missaris, Vasas, & McCloskey, 1987; Commissaris, McCloskey, Damian, Brown, Barraco, & Altman, 1990; McCloskey, Paul, & Commissaris, 1987), pentobarbital (Commissaris, Vasas, & McCloskey, 1987), and secobarbi-tal (Commissaris, Vasas, & McCloskey, 1987), all of which are GABAergic.
Multiple Schedules of Reinforcement and Punishment A multiple schedule involves two schedules of reinforcements that are alter-nately placed in effect, each signaled by a different stimulus. Geller and Sei-fter (1960) used the multiple schedule to study anti-punishment effects by programming a simultaneous schedule of punishment and reinforcement in one component and comparing behavior to a second component with no pun-ishment schedule. The advantage the multiple schedule has over the simple schedule is that unpunished and punished behavior can be examined readily within a single session for each subject. In Geller and Seifter's (1960) study, the bar-pressing of rats was placed under a multiple variable interval (VI) 2’-- continuous reinforcement (CRF) plus shock schedule. In the VI 2’ com-ponent, the fi rst response after an average of 2 minutes elapsed, produced a reinforcer. Under the CRF plus shock component, every response produced a food pellet and a shock, thereby suppressing lever-pressing in that compo-nent only. A single dose (120 mg/kg) of meprobamate increased the number of lever-presses in the CRF component by an order of magnitude (control mean=3.22, meprobamate mean=33.67). Lever-press increases were not ob-served in the VI (unpunished) component. Phenobarbital and pentobarbital dose-dependently increased the number of punished, but not unpunished, responses as well. Promazine (an antipsychotic) and d-amphetamine de-creased the number of punished responses, again showing anti-punishment effects that are specifi c to anxiolytic drugs. In a later study, the Geller-Seifter procedure was used to show that reserpine, a depressant that acts by deplet-ing norepinephrine and serotonin (5-HT), revealed anti-punishment effects, but morphine, an opiate, did not (Geller, Bachman, & Seifter, 1963.) The lack of effect to morphine is important because it indicates that anti-punishment effects are not related to analgesia.
Other studies have replicated anti-punishment effects with multiple schedules across a variety of parameters, including the type of aversive stim-uli. Several studies found that chlordiazepoxide and pentobarbital increased punished responding when timeout (Branch, Nicholson, and Dworkin, 1977; McMillan, 1967; van Haaren and Anderson, 1997) and pressurized air (Speal-man, 1979) were used as aversive stimuli. Anti-punishment effects revealed by the Geller-Seifter procedure have also been found with other GABAergic compounds, including amobarbital (e.g., Davidson and Cook, 1969; Morse, 1964), chlordiazepoxide (e.g.,Glowa and Barrett, 1976; Koob, et al., 1986; Sethy and Winter, 1972), ethanol (e.g., Koob, et al., 1986; Glowa and Barrett, 1976), phenobarbital (e.g., Commissaris, Vasas, and McCloskey, 1988), and secobarbital (e.g., Commissaris, Vasas, and McCloskey, 1988). These effects have been observed also with chlorpromazine (Dinsmoor and Lyon, 1961), an anti-psychotic dopamine antagonist, though this drug has inconsistent fi nd-ings associated with anti-punishment (cf. Morse, 1964; Pollard & Howard, 1979, for example).
Concurrent Schedules of Reinforcement The concurrent variable interval variable interval (conc VI VI) schedule of rein-forcement has been used to examine anti-punishment effects in the context of choice behavior. Here, responding under two simultaneously available sched-ules of reinforcement produces reinforcers that are delivered at an overall predictable rate, although the moment-to-moment deliveries vary. Behavior allocation between the two alternatives matches the relative reinforcement rates (e.g., Herrnstein, 1961). In other words, if twice as many reinforcers are delivered from alternative A than alternative B, twice as much behavior will be allocated to alternative A than alternative B. Aversive stimuli, such as shock, have been used with this procedure in an effort to examine how punishment affects behavior allocation (e.g., Wojnicki & Barrett, 1993). With humans, less noxious stimuli, such as noise (e.g., Katz, 1973) or point and money loss (e.g., Bradshaw, Szabadi, & Bevan, 1979; Critchfi eld, Paletz, & MacAleese, 2003; deVillers, 1980; Rasmussen & Newland, 2006) have been used. The concur-rent schedule is useful in studying punishment effects because if behavior under one alternative is punished, the behavior of the unpunished alternative is available for allocation. Behavior, then, can be “captured” under the unpun-ished alternative. When punishment is superimposed on simple or multiple schedules of reinforcement, the probability of interacting with the key, bar, or whatever is associated with punishment is lowered and alternative behavior is more probable. The precise measurement and recording of “other” behavior that occurs in place of punished behavior (e.g., cage exploration, grooming, sleeping or sniffi ng) is diffi cult. The only dependent variable an experimenter has to examine in terms of free-operant behavior is the very few responses that occur toward the key or lever. With concurrent schedules, the environ-ment is arranged in such a way that behavior toward the unpunished alterna-tive can occur (i.e., the rat, for example, can still lever-press for reinforcers instead of grooming or exploring.) Within the context of studying anti-punish-ment effects, it is important because it is possible to examine concurrently an unpunished and punished condition—the putative “litmus test” for specifi city.
Anti-punishment effects have been explored using concurrent schedules. Wojnicki and Barrett (1993) conducted a study with a conc VI VI schedule that included a superimposed schedule of punishment on one alternative. They examined behavior in the punished and unpunished alternatives under vary-ing doses of chlordiazepoxide, buspirone, and d-amphetamine. Responses under the punished component increased in a dose-dependent fashion un-der chlordiazepoxide and buspirone, but not under amphetamine. Behavior under the unpunished component decreased slightly at higher doses for the two anxiolytics, and decreased under amphetamine. Sepinwall, Grodsky, and Cook (1978) found similar results under concurrent schedules with squirrel monkeys and the drugs diazepam and chlordiazepoxide. Anti-punishment ef-fects have been demonstrated under concurrent schedule with humans with the drugs diazepam (Carlton et al., 1981) and ethanol (Rasmussen & New-land, 2006).
ANTI-CONFLICT
As mentioned previously, anti-confl ict studies involve examining behavior under a simultaneously appetitive (approach) and aversive (avoidance) con-dition, broadly defi ned. An anxiolytic drug functions to increase avoidance behavior. The following procedures represent anti-confl ict studies, and are meant to be an illustrative, not exhaustive, refl ection of the extant literature. In the elevated plus maze two long planks bisect each other at 90 degree angles. The bisection is usually enclosed as a small compartment. Two of the four protruding arms are enclosed within walls and the other two are ex-posed. A rodent, usually a rat, is placed in the center compartment and the arms are available for exploration. Under non-drug conditions rats tend to traverse the enclosed arms more often and to a greater extent than the ex-posed arms (Montgomery, 1955). Several studies have shown that GABAer-gic drugs, such as diazepam (e.g., Balfour, Graham, & Vale, 1986; Moser, 1989) and ethanol (e.g., Lister, 1988) selectively increase open-arm explora-tion and time spent in the exposed arm. Some atypical anxiolytics (5-HT ago-nists), such as ipsapirone, PAPP, (-)MDL 72832, and (+)MDL 72832 also in-crease open arm entries (Moser, 1989). However, buspirone has been found to decrease open-arm entries (Moser, 1989; Pellow, Johnston, & File, 1987). Recall in the light-dark task, under non-drug conditions, a rodent may allocate more time in the dark compartment of a light-dark box and avoid the light compartment. Benzodiazepines may increase entries and time spent in the light compartment. For example, Onaivi and Martin (1989) reported that under control conditions, mice were more active in a dark chamber, even when it was smaller or equal in size to a light chamber. Some doses of diazepam and buspirone, but not amphetamine and morphine, signifi cantly increased the ex-ploratory activity in the light cell, and did not increase activity in the dark cell. These fi ndings have been replicated with diazepam (Costall et al., 1987) and found with ethanol (Belzung, Misslin, and Vogel, 1988; Costall et al., 1987). Other Procedures Anti-confl ict effects also have been reported in open fi eld activity, in which food (the appetitive stimulus) is placed in an open fi eld. Here, the putative competing forces are said to be food and neophobia (the aversive stimulus, especially if the environment is brightly lit). Bodnoff, Suranyi-Cadotte, Quiron, and Meaney (1989) reported decreased latencies to contact with food when rats were chronically administered a variety of drugs, including diazepam, fl uoxetine, buspirone, mainserin, and gepirone. Interestingly, the decreases in latencies were selective to anxiolytics (diazepam and buspirone) and anti-depressants (fl uoxetine, mainserin, and gepirone). However, these fi ndings are limited because there was no test for the selectivity of aversive versus ap-petitive conditions, since the inferred drives were untestable. Moreover, only a single dose of each drug was used, so complete dose-response profi les for each drug were not obtained. The six-foot alley is another confl ict procedure in which anti-confl ict effects have been demonstrated. Grossman and Miller (1961) conducted a study in which food was placed at the end of a six-foot alley. Each foot was then as-sociated with a progressively larger increase in shock intensity, such that no shock was correlated with the fi rst foot and the largest intensity shock (180 volts) was associated with the sixth foot. Traversing speed within each one-foot length was recorded under both shocked and non-shocked conditions. Chlorpromazine and ethanol increased speed in each one-foot component under shocked, but not non-shocked conditions. Studies that examine “taboo” behavior with humans may qualify as anti- confl ict studies. One procedure is the slide rate measure in which slides with sexual, aggressive, or neutral themes are presented to participants. Under ethanol, participants spend more time than under non-drug conditions watch-ing those of a sexual or aggressive, but not neutral, nature than under non-drug conditions (Kallmen & Gustafson, 1998). Here, the putative competing forces are said to be the “inherently” reinforcing nature of sex or violence versus the socially punishing contingencies of viewing pornographic or violent images, which the subjects supposedly bring with them to the experiment.
Aggression in humans is another taboo behavior that may apply to anti- confl ict. Dougherty, Cherek, and Bennett (1996) examined the number of ag-gressive responses by women that occurred under different doses of ethanol. Participants could push a button to earn points exchangeable for money. They could also push another button to take points from another “person,” which was interpreted by the authors as an aggressive response. To induce aggres-sive responses, subjects were told that every time their counters lost a point it was because a person in another room took the point. The number of aggres-sive responses increased with ethanol dose, and under the highest dose of ethanol (1 g/kg) aggressive responses doubled. The number of point-earning responses did not increase under ethanol. Although there was no punishment condition in this study to suppress behavior, the fi ndings may qualify as anti-confl ict because aggressive responding occurred at lower rates initially. This fi nding might imply that extra-experimental contingencies may serve to pun-ish behavior, and the participants brought this history with them to the labora-tory. Similar results have been reported with male participants (see Cherek & Steinberg, 1987; Cherek, Steinberg, and Manno, 1985; Cherek, Steinberg, & Vines, 1984) and seem to be specifi c to ethanol, as d-amphetamine (Cherek, Steinberg, & Kelly, 1986), caffeine (Cherek, Steinberg, & Brauchi, 1983), and nicotine (Cherek, 1981) produced behavioral changes that were not specifi c to aggressive responding.
COMPARING ANTI-PUNISHMENT
AND ANTI-CONFLICT PROCEDURES
Anti-confl ict is similar to anti-punishment in that behavior that occurs at a low frequency under non-drug conditions selectively emerges during drug con-ditions. The difference between confl ict and punishment, however, rests on several important details. In confl ict procedures the reinforcement contingen-cies are poorly defi ned-- behavior often is discussed as the resolution of “two opposing motivational forces” (p. 443; Commissaris, 1993). For example, in the light-dark task, the appetitive condition is defi ned as the animal’s drive to explore and the aversive condition as its innate tendency to avoid light (Commissaris, 1993). One can identify the problem of subjectivity and an-thropomorphism that may arise from defi ning the conditions as such. More important, other problems are evident in using such characterizations, such as assumptions of a behavioral history, (more specifi cally, a history of punish-ment) and the inability to observe stability in behavior. Punishment proce-dures, conversely, rely on simple and easily interpretable characterization of punished and unpunished behavior, though they certainly could be framed within a “confl ict” defi nition. A second difference between the two types of procedures is that anti-con- fl ict procedures tend to rely on more naturalistic behavioral repertoires of the subjects studied, for example, approach to darkness, exploring, aggression, etc. Anti-punishment studies (with the exception of licking) involve arbitrary responses with aversive stimuli that are unlikely to be a part of an organism’s natural environment (e.g., shock). Hence, anti-confl ict procedures can be ar-gued to have more ecological validity and generalization to the real world, while anti-punishment procedures can be argued to result in more precise defi nition of the variables and tighter control of the experimental environment, resulting in more internal validity. Consideration of both types of procedures is essential for a fuller characterization of behavior under anxiolytic drugs, but to categorize anti-punishment and anti-confl ict as interchangeable is inac-curate.
A third difference between the sets of procedures is that under punish- ment behavior is characterized fi rst within the unpunished (“approach”) con-ditions (i.e., baseline), and then the aversive stimulus is delivered as a pun-isher, allowing a situation in which specifi city is easily interpretable. Confl ict assumes that some conditions are aversive and others are appetitive, based on the animals’ behavior upon fi rst experience, but without showing this in a functional manner. In some procedures (for example, open fi eld activity), there is no built-in test for specifi city of appetitive versus aversive conditions.
A fourth difference, and probably the most compelling, is that anti-punish- ment involves behavior that specifi cally has been punished, and anti-confl ict is one that likely involves avoidance, or negative reinforcement—these two phenomena have different behavioral and pharmacological mechanisms. Be-haviorally, punishment reduces behavior by the presentation of an aversive stimulus and negative reinforcement increases behavior by the removal of an aversive stimulus. As discussed, and as shown in Table 1, pharmacologi-cally, anti-punishment effects have been identifi ed specifi cally with GABAer-gic drugs, namely benzodiazepines, barbiturates, and alcohol and less so with serotonergic ones; anti-confl ict effects have been related to GABAergic compounds, a much wider variety of 5-HT agonists (second-generation anxio-lytics), dopaminergic agonists, and some other drugs that do not necessarily have anxiolytic properties (such as 5-HT antagonists and nicotine). Studies on the neurochemical basis of negative reinforcement using pro- cedures other than anti-confl ict suggest that a wider range of neurotransmitter systems seems to be involved in negative reinforcement. Shock-maintained behavior, in which postponement of shock maintains the response that pro-duces the shock postponement, is an example. Two-factor theorists argue that the maintenance of the response involves escape from anxiety or fear as- sociated with time nearing the delivery of the aversive event (Mowrer, 1947). (The debate on one- and two-factor theories of avoidance is captured nicely in an amalgam of commentaries of the May 2001 issue of the Journal of the Experimental Analysis of Behavior.) In line with two-factor theory, then, drugs that are said to reduce anxiety should result in a decrease in shock-maintained avoidance responses (without disabling motor function). The anx-iolytics buspirone (Galizio, Hale, Librorio, & Miller, 1993), diazepam (Kuri-bara, 1978), pentobarbital (Kuribara, 1978), and ethanol (Galizio, Perone, & Spencer, 1986) have been shown to decrease shock-maintained avoidance responses. However, other anxiolytics, such as chlordiazepoxide and 8-OH-DPAT (Galizio, Hale, Librorio, & Miller, 1993) have been shown to increase shock-maintained avoidance responses. In addition, some muscarinic ago-nists and dopamine antagonists, also known as anti-psychotics, have re-duced conditioned shock avoidance responses in discrete trials procedures (Shannon, et al, 1999). Moderate doses of the anti-psychotic chlorpromazine have been shown to decrease also free operant shock-maintained avoidance responses (Galizio, Journey, Royal, & Welker, 1909; van Haaren, & Zarcone, 1994), though avoidance responses increased at low doses. It seems, then, that when it comes to shock-maintained responses, the role of anxiety, as referenced by response to anxiolytics, is not central or straightforward.
Anxiolytics that affect other event-maintaining responses (in addition to shock) have been examined. Responses that produce signaled timeout from avoidance have been shown to decrease with administration of buspirone. However, opiate agonists, including morphine, methadone, fentanyl, and U50-488 also decrease avoidance responses (Galizio, Robinson, & Ordronneau, 1994), though the opiates also increased shock-maintained avoidance, but to a lesser extent than timeout. These data, combined with the puzzling data on the behavioral pharmacology of shock-maintained avoidance suggests that procedural variants such as the event itself that maintains avoidance (e.g., timeout, shock avoidance, escape from a conditioned stimulus), rather than anxiety or fear, may be a play a key role in understanding the maintenance of avoidance. In other words, if anxiety was the internal stimulus that was reduced with anxiolytic drugs, the same anxiolytic drugs should reduce avoid-ance responses in an avoidance task, despite what type of aversive event maintains the behavior. Moreover, only drugs that have been shown to be anxiolytic should decrease avoidance responses; anti-psychotics and mus-carinic agonists should not. It may well be the case that the type of event that maintains the response in avoidance is a non-trivial matter in terms of exam-ining drug effects. Consider Verhave’s (1962) view of timeout from avoidance, which suggests that timeout may function more as a positive reinforcer (addi-tion of a safety stimulus) as opposed to a negative reinforcer. Views such as these (as opposed to appealing to anxiety as explanation) are more likely to lead to better formulated hypotheses about how drugs will affect a behavior, and will likely lead to a clearer understanding of drug-behavior interactions. It seems then, that the pharmacological mechanism of negative reinforce- ment is not straightforward. The behavioral and pharmacological mechanisms of negative reinforcement may depend on such parameters as the event main-taining the behavior, event-independent factors (e.g., rate dependence—see Galizio and Allen, 1991), as well as some possible species differences (which may make members of one species more sensitive to the aversive or re-inforcing properties of particular events, e.g., light). These parameters may involve different and multiple neurochemical substrates that when taken to-gether would result in a nebulous and confusing picture, but when examined independently, may provide a clearer picture of the pharmacological basis of negative reinforcement.
It seems clear, though, that behavior maintained by negative reinforcement is pharmacologically (and behaviorally) different from behavior maintained by punishment. Anti-punishment effects hold to GABA related drugs, which likely function to make behavior less sensitive the punishing contingency. The neg-ative reinforcement nature of anti-confl ict procedures holds to a larger range of drugs that may not fi t neatly into the anxiolytic pharmacological class. In-deed, the picture with anti-confl ict procedures is much more complex than that of anti-punishment. Moreover, the common internal mechanism of anxi-ety reduction as an explanation for the drug’s effi cacy in releasing suppressed behavior does not seem to be supported pharmacologically between the two categories of procedures. Indeed if anxiety reduction was the mechanism for behavioral change with anxiolytic drugs, there would be a greater overlap in drugs that affect behavior under anti-punishment and anti-confl ict procedures. Because anti-punishment and anti-confl ict procedures are distinct in terms of the behavioral mechanisms and pharmacological mechanisms involved, it makes little sense to classify them in an interchangeable manner.
Recently, anti-confl ict, and some anti-punishment procedures have been used to disentangle 5-HT receptor subtypes (e.g., Dhonnchadha, Hascoet, Jolliet, & Bourin, 2003), and to compare anti-confl ict/anti-punishment effects to discriminative properties of the drugs (e.g., McMillan, Li, & Hardwick, 1997; Pattij, Hijzen, Maes, & Olivier, 2000.) The latter represents a step in the right direction for behavior analysis, because the question that arises from compar-ing discriminative and anti-confl ict/anti-punishment properties is whether the stimulus properties of a drug are necessary to disrupt the behavior-releasing effects of a drug. In other words, perhaps the behavioral mechanism for anti-confl ict or anti-punishment has to do with yet another behavioral mechanism, in addition to event-dependent (or event-independent) effects--the internal stimulus properties that may, or may not exert infl uence on punished behavior or negatively reinforced behavior. Hypotheses of the relation between the dis-criminative properties and the behavior-releasing effects of anxiolytic drugs suggest that the neurochemical substrate underlying the two mechanisms may have some similarity (Pattij, et al, 2000). Answers to these questions may become clear as more studies are conducted.
Balfour, D., Graham, C., & Vale, A. (1986). Studies on the possible role of brain 5-HT systems and adrenocortical activity in behavioral responses to nicotine and diaz-epam in an elevated plus maze. Psychopharmacology, 90, 528-532.
Barrett, J., Brady, L., & Witkin, J. (1985). Behavioral studies with anxiolytic drugs. I. Interactions of the benzodiazepine antagonist Ro 15-1788 with chlordiazepoxide, pentobarbital and ethanol. Journal of Pharmacology & Experimental Therapeu-tics, 233, Belzung, C., Misslin, R., & Vogel, E. (1988). The benzodiazepine receptor inverse CCM and Ro 15-3505 both reverse the anxiolytic effects of ethanol in mice. Life Bodnoff, S., Suranyi-Cadotte, B., Quiron, R., & Meaney, M. (1989). A comparison of the effects of diazepam versus several typical and atypical anti-depressant drugs in an animal model of anxiety. Psychopharmacology, 97, 277-279.
Bradshaw, C., Szabadi, E., & Bevan, P. (1979). The effect of punishment on free-oper- ant choice behavior in humans. Journal of the Experimental Analysis of Behavior, 31, 71-81.
Branch, M., Nicholson, G., & Dworkin, S. (1977). Punishment-specifi c effects of pento- barbital: dependency on the type of punisher. Journal of the Experimental Analy-sis of Behavior, 28, 285-293.
Brocco, M.J., & Koek, W., Degryse, A.-D., & Colpaert, F.C. (1990). Comparative stud- ies on the anti-punishment effects of chlordiazepoxide, buspirone, and ritaserin in the pigeon, Geller-Seifter and Vogel confl ict procedures. Behavioural Pharmacol-ogy, 1, 403-418.
Carlton, P., Siegel, J., Murphee, H., & Cook, L. (1981). Effects of diazepam on operant behavior in man. Psychopharmacology, 73, 314-317.
Cherek, D.R.(1981). Effects of smoking different doses of nicotine on human aggres- sive behavior. Psychopharmacology, 75, 339-345.
Cherek, D.R., & Steinberg, J.L.(1987). Effects of drugs on human aggressive respond- ing. In G.D. Burrows & J.S. Werry (Eds.) Advances in human psychopharmacol-ogy: A research annual, Vol. 4., US: Elsevier Science/JAI Press, 239-290.
Cherek, D.R., Steinberg, J.L., & Brauchi, J.T. (1983). Effects of caffeine on human ag- gressive behavior. Psychiatry Research, 8, 137-145.
Cherek, D. R. Steinberg, J.L., & Kelly, T.H.(1986). Effects of d-amphetamine on human aggressive behavior. Psychopharmacology, 88, 381-386.
Cherek, D., Steinberg, J., & Manno, B. (1985). Effects of ethanol on human aggressive behavior. Journal of Studies on Ethanol, 46, 321-327.
Cherek, D., Steinberg, J., & Vines, R. (1984). Low doses of ethanol affect human ag- gressive responding. Biological Psychiatry, 19, 263-267.
Commissaris, R. (1993). Confl ict behaviors as animal models for the study of anxiety. In F. van Haaren, (Ed.) Research methods in behavioral pharmacology. Amster-dam: Elsevier Science Publishers.
Commissaris, R.L., Harrington, G.M., & Altman, H.J. (1990). Benzodiazepine anti-con- fl ict effects in Maudsley Reactive and Non-Reactive Strains. Psychopharmacol-ogy, 100, 287-292.
Commissaris, R.L., McCloskey, T.C., Damian, G.M., Brown, B.D., Barraco, R.A., & Altman, H.J. (1990). Antagonism of the anti-confl ict effects of Phenobarbital, but not diazepam, by the A-1 adenosine agonist I-PIA. Psychopharmacology, 102, 283-290.
Commissaris, R.L., & Rech, R.H. (1982). Interactions of metergoline with diazepam, quipazine, and hallucinogenic drugs on a confl ict behavior in the rat. Psychophar-macology,76, 282-285.
Commissaris, R., Vasas, R., & McCloskey, T. (1988). Convulsing versus typical bar- biturates: Effects on confl ict behavior in the rat. Pharmacology, Biochemistry, & Behavior, 29, 631-634.
Costall, B., Kelly, M., & Naylor. (1987). The anxiolytic and anxiogenic actions of etha- nol in a mouse model. Journal of Pharmacology, 40, 197-202.
Critchfi eld, T.S., Paletz, E.M., & MacAleese, K.R. (2003). Punishment in human choice: Direct or competitive suppression? Journal of the Experimental Analysis of Behavior, 80, 1-27.
Davidson, A., & Cook, L. (1969). Effects of combined treatment with trifl uoperazine- HCl and amobarbital on punished behavior in rats. Psychopharmacologia, 15, 159-168.
deVillers, P. (1980). Toward a quantitative theory of punishment. Journal of the Experi- mental Analysis of Behavior, 33, 15-25.
Dhonnchadha, B.A.N., Hasscoet, M., Jolliet, P., & Bourin, M. (2003). Evidence for a 5-HT2A receptor mode of action in the anxiolytic properties of DOI in mice. Behav-ioural Brain Research, 147, 175-184. Dinsmoor, J., & Lyon, D. (1961). The selective action of chlorpromazine of behavior suppressed by punishment. Psychopharmacologia, 2, 465-460.
Dougherty, D., Cherek, D., & Bennett, R. (1996). The effects of ethanol on the aggres- sive responding of women. Journal of Studies on Ethanol, 23, 178-186.
Dworkin, S.I., Bimle, C., & Miyauchi, T. (1989). Differential effects of pentobarbital and cocaine on punished and unpunished responding. Journal of the Experimental Analysis of Behavior, 51, 173-184.
Estes, W.K., & Skinner, B.F. (1941). Some quantitative properties of anxiety. Journal of Experimental Psychology, 29, 390-400.
Fontana, D.J., Carbary, T.J., & Commissaris, R.L. (1989). Effects of acute and chronic anti-panic drug administration on confl ict behavior in the rat. Psychopharmacol-ogy, 98, 157-162.
Galizio, M., Allen, A.R. (1991). Variable-ratio schedules of timeout from avoidance: Effects of d-amphetamine and morphine. Journal of the Experimental Analysis of Behavior, 56, 193-203.
Galizio, M., Hale, K. L., & Liborio, M.O. (1993). Variable-ratio schedules of timeout from avoidance: Effects of anxiolytic drugs. Behavioural Pharmacology, 4, 487-493.
Galizio, M., Journey, J.W., Royal, S.A. (1990). Variable-interval schedules of timeout from avoidance: Effects of anxiolytic and antipsychotic drugs in rats. Pharmacol-ogy, Biochemistry & Behavior, 37, 235-238 Galizio, M., Perone, M., Spencer, B.A. (1986) Variable interval schedules of timeout from avoidance: Effects of ethanol, naltrexone, and CGS 8216. Pharmacology, Biochemistry & Behavior, 25, 439-448 Galizio, M., Robinson, E. G. & Ordronneau, C. (1994). Opioid drugs and timeout from avoidance. Behavioural Pharmacology, 5, 125-130 Geller, I., Bachman, E, & Seifter, J. (1963). Effects of reserpine and morphine on be- havior suppressed by punishment. Life Sciences, 4, 226-231.
Geller, I., & Seifter, J. (1960). The effects of meprobamate, barbiturates, d-amphet- amine, and promazine on experimentally-induced confl ict in the rat. Psychophar-macologia, 1, 482-492.
Glowa, J.R., & Barrett, J.E. (1976). Effects of ethanol on punished and unpunished responding of squirrel monkeys. Pharmacology, Biochemistry, and Behavior, 4, 169-173.
Gray, J.A. (1988.) Behavioural and neural-system analyses of the actions of anxiolytic drugs. Pharmacology, Biochemistry, and Behavior BB, 29, 767-769.
Grossman, S., & Miller, N. (1961). Control for stimulus change in the evaluation of ethanol and chlorpromazine as fear-reducing drugs. Psychopharmacologia, 2, 342-351.
Herrnstein, R. (1961). Relative and absolute strength of response as a function of rein- forcement. Journal of the Experimental Analysis of Behavior, 4, 267-272.
Kallmen, H., & Gustafson, R. (1998). Ethanol and disinhibition. European Addiction Katz, R. (1973). Effects of punishment in an alternative response context as a function of relative reinforcement rate. Psychological Record, 23, 65-74.
Kennett, G.A., Trail, B., & Bright, F. (1998). Anxiolytic-like actions of BW 723C86 in the rat Vogel confl ict test are 5-HT2B receptor mediated. Neuropharmacology, 37, 1603-1610. Kilts, C.D., Commissaris, R.L., & Rech, R.H. (1981). Comparison of anti-confl ict drug effects in three experimental animal models of anxiety. Psychopharmacology, 74, 290-296.
Koek, W., & Colpaert, F.C. (1991). Use of a confl ict procedure in pigeons to characterize anxiolytic drug activity: Evaluation of NMDA antagonists. Life Sciences, 49, 37-42.
Koob, G., Braestrup, C., & Britton, K. (1986). The effects of FG 7142 and Ro 15-1788 on the release of punished responding produced by chlordiazepoxide and ethanol in the rat. Psychopharmacologia, 90, 173-178.
Kuribara, H. (1978.) Psychotropic drugs and Sidman avoidance in rats: IRT distribution changes. Pharmacology, Biochemistry, & Behavior, 8, 537-542.
Lerner, T., Feldon, J., & Myslobodsky, M.S. (1986). Amphetamine potentiation of anti- confl ict action of chlordiazepoxide. Pharmacology, Biochemistry, & Behavior, 24, 241-246.
Liljequist, S., & Engel, J. (1984). The effects of GABA and benzodiazepine receptor antagonists on the anti-confl ict actions of diazepam or ethanol. Pharmacology, Biochemistry & Behavior, 21, 521-5.
Lister, R. (1988). Interactions of three benzodiazepines inverse agonists with ethanol in a plus-maze test of anxiety. Pharmacology, Biochemistry and Behavior, 30, 701-706.
MacCorquodale, K., & Meehl, P.E. (1948) On a distinction between hypothetical con- structs and intervening variables. Psychological Review, 55, 95-107.
Mansbach, R.S., Harrod, C., Hoffman, S.M., Nader, M.A., Lei, Z., Witkin, M., & Barrett, J.E. 1988). Behavioral studies with anxiolytic drugs. V. Behavioral and in vivo neurochemical analyses in pigeons of drugs that increase punished responding. The Journal of Pharmacology and Experimental Therapeutics, 246, 114-120.
Masserman, J., & Yum, K. (1946). An analysis of the infl uence of ethanol on experi- mental neurosis in cats, Psychosomatic Medicine, 8, 36-52.
McCloskey, T.C., Paul, B.K., & Commissaris. R.L. (1987). Buspirone effects in an ani- mal confl ict procedure: Comparison with diazepam and phenobarbital. Pharma-cology, Biochemistry, & Behavior, 27, 171-175.
McMillan, D. (1967). A comparison of the punishing effects of response-produced shock and response-produced timeout. Journal of the Experimental Analysis of Behavior, 10, 439-449.
McMillan, D., Li, & Hardwick. (1997). Drug discrimination under a concurrent fi xed- interval fi xed-interval schedule. Journal of the Experimental Analysis of Behav-ior,68, 193-217.
Montgomery, K. (1955). The relation between fear induced by novel stimulation and exploratory behavior. Journal of Comparative and Physiological Psych, 48, 254-260.
Morse, W. (1964). Effect of amobarbital and chlorpromazine on punished behavior in the pigeon. Psychopharmcologia, 6, 286-294.
Moser, P. (1989). An evaluation of the elevated plus-maze test using the novel anxio- lytic buspirone. Psychopharmacology, 99, 48-53.
Mowrer, O.H. (1947). On the dual nature of learning of learning-- a reinterpretation of "conditioning" and "problem solving". Harvard Educational Review, 17, 102-148.
Onaivi, E., & Martin, B. (1989). Neuropharmacological and physiological validation of a computer-controlled two-compartment black and white box for the assess-ment of anxiety. Prog. Neuropsychopharmacological and Biological Psychiatry, 13, 963-976.
Pattij, T., Hijzen, T.H., & Gommans, J. (2000). Effects of drug discrimination history on anti-punishment properties of chlordiazepoxide in rats. Pharmacology, Biochem-istry & Behavior, 67, 621-627. Pellow, S., Johnston, A., & File, S. (1987). Selective agonists and antagonists for 5- hydroxytryptamine receptor subtypes and interactions with yohimbine and FG 7142 using the elevated-plus maze in the rat. Journal of Pharmacology, 39, 917-928.
Pollard, G., & Howard, J. (1979). The Geller-Seifter confl ict paradigm with incremental shock. Psychopharmacology, 62, 117-121.
Rasmussen, E.B., & Newland, M.C. (2005). Anti-punishment effects of ethanol in hu- mans quantifi ed by the generalized matching relation. Manuscript under review.
Schefke, D.M., Fontana, D.J., & Commissaris, R.L. (1989). Anti-confl ict of buspirone following acute versus chronic treatment. Psychopharmacology, 99, 427-429.
Sepinwall, J., Grodsky, F., & Cook, L. (1978). Confl ict behavior in the squirrel mon- key: Effects of chlordiazepoxide, diazepam and n-desmethyldiazepam. Journal of Pharm and Exper Therapeutics, 204, 88-102.
Sethy, V., & Winter, J. (1972). Effects of yohimbine and mescaline on punished behav- ior in the rat. Psychopharmacologia, 23, 160-166.
Spealman, R. (1979). Comparison of drug effects on responding punished by pres- surized air or electric shock delivery in squirrel monkeys: pentobarbital, chlordiaz-epoxide, d-amphetamine and cocaine. Journal of Pharmacology & Experimental Therapeutics. Stefanski, R., Palejko, W., Kostowski, W., & Plaznik, A. (1992). The comparison of benzodiazepine derivatives and serotonergic agonists and antagonists in two ani-mal models of anxiety. Neuropharmacology, 31, 1251-1258, Thiebot, M., Soubrie, P., & Simon, P. (1985). Is delay reward mediated by shock avoid- ance behavior a critical target for anti-punishment effects of diazepam in rats? Psychopharmacology, 87, 473-479.
van Haaren, F., & Anderson, K. (1997). Effects of chlordiazepoxide, buspirone and cocaine on behavior suppressed by timeout. Behavioural Pharmacology, 8, 174-182.
van Haaren, F., & Zarcone, T. J. (1994). Effects of chlordiazepoxide and cocaine on concurrent food and avoidance-of-timeout schedules. Journal of the Experimental Analysis of Verhave, T. (1962). The functional properties of a timeout from avoidance schedule. Journal of the Experimental Analysis of Behavior, 5, 391-422.
Vogel, J., Beer, B., & Clody, D. (1971). A simple and reliable confl ict procedure for test- ing anti-anxiety agents. Psychopharmacologia, 21, 1-7.
Vogel, R. (1980). Attenuation of the effects of punishment by ethanol: Comparisons with chlordiazepoxide. Psychopharmacology, 71, 123-129.
Waddington, J.L., & Olley, J.E. (1977). Dissociation of the anti-punishment activities of chlordiazepoxide and atropine using two heterogeneous passive avoidance tasks. Psychopharmacology, 52, 93-96.
Witkin, J.M., Mansbach, R.S., Barrett, J.E., Bolger, J.E., Skolnick, P., & Weissman, B. (1987). Behavioral studies with anxiolytic drugs. IV. Serotonergic involvement in the effects of buspirone on punished behavior of pigeons. The Journal of Pharma-cology and Experimental Therapeutics, 243, 970-977.
Wojnicki, F., & Barrett, J. (1993). Anti-confl ict effects of buspirone and chlordiazepox- ide in pigeons under a concurrent schedule of punishment and a changeover response. Psychopharmacology, 112, 26-33.

Source: http://wwws.isu.edu/psych/Articles/Rasmussen/Rasmussen%20(2006)..pdf