Back to book TOC
|(Search book contents)||(Search entire ASNet)|
Conditioned Reinforcement as a Measure of
the Rewarding Properties of Drugs
W. Marvin Davis and Stanley G. Smith*
Department of Pharmacology
School of Pharmacy
The University of Mississippi
University, Mississippi 38677
|A review is presented of the development and application of a conditioned reinforcement measure for evaluating drug-based primary reinforcement. This procedure has employed both contingent and noncontingent association of the initially neutral stimulus with repeated doses of the drug reinforcer. Evidence for primary reinforcement by doses of drug is later demonstrated by the behavior of a subject when the stimulus is presented contingent to lever-pressing responses. This method allowed testing of potential antagonist drugs for an ability to oppose or cancel the reinforcing action of agents such as dexamphetamine, morphine, and ethanol. Inhibitors of critical enzymes in catecholamine biosynthetic pathways (e.g., tyrosine hydroxylase or dopamine ß-hydroxylase) were found to oppose the reinforcing action of such classical drugs of abuse. The ability of conditioned reinforcement tests to avoid the factor of performance deficits induced by many potential antagonist drugs constitutes a major advantage over a more direct evaluation of drugs for possible interactions with primary reinforcing actions of abuse agents.|
At the beginning of the 1970s, we began investigations of the role that central nervous system (CNS) biogenic amines play in the production of drug-based reinforcement. Our initial efforts centered on the development of experimental protocols to ensure the validity and reliability of planned research on opiate- and amphetamine-based reinforcement. We perceived that previous self-administration research had not ruled out completely the possibility that such drugs were only pseudo-reinforcers, functioning as performance facilitators. That is, previous studies had not controlled for the possibility that drug delivery might merely have increased the general baseline of motor behavior, thus mimicking reinforcement. Also, we soon became aware that pharmacological tools that we wished to employ for the study of CNS adrenergic functions could alter a data baseline via inhibition of motor function, independently of their effects in the central processes related to
*Present address: Choctaw Community Mental Health Center, Philadelphia, Mississippi drug-based reinforcement. Again, this raised the possibility that data generated using such agents might reflect pharmacological side-effects. Therefore, it seemed imperative to find methods that could validly demonstrate drug-based reinforcement without performance artifacts and that would allow the manipulation of the CNS neurotransmitter systems by means of pharmacological tools without unwanted actions such as motor inhibition, motor disinhibition, or the introduction of nonreinforcement factors into the reinforcement measure.
The area of intracranial self-stimulation research seemed both to suffer parallel problems and to provide a means for circumventing such problems in valid measurement of reinforcement associated with nonclassical sources of reward (Liebman, 1983; Valenstein, 1964). Stein (1958) reported that electrical stimulation of the brain in rats had the capacity to impart secondary reinforcing properties to an originally neutral stimulus. This was accomplished by means of four daily pairing sessions with 100 presentations of a 1.0-second tone given concurrently with a 0.5-second train of electrical stimuli delivered to the brain. Three days following the pairings, a 1-hour test session was conducted in which the rats had access to two levers, and only one lever activated the same 1.0-second tone as in the pairings. This lever had been nonpreferred in a baseline session before the pairings. The no-tone lever only measured general motor activity. Test results showed that the lever-pressing on the tone-producing lever had increased significantly in subjects for whom the brain stimulations were subsequently proven (by giving an opportunity to lever-press for the brain stimulation) to have been positively reinforcing. Subjects whose electrodes were found to give a neutral response (i.e., lacking positive reinforcement) showed neither a change in lever preferences nor an increase in rate of lever-pressing subsequent to tone-stimulation pairings.
We saw Stein's experimental design to be applicable to our aims, substituting small intravenous (i.v.) drug doses for electrical brain stimulation. It provided a means of testing for reinforcement, allowing simultaneous recording of data to control for drug-produced changes in motor activity and thus avoiding performance interpretations. It permitted testing for the reinforcer's stimulus control in the acquisition of a newly learned response, further weakening the performance interpretation. It allowed testing for an interaction between the primary reinforcer and a pharmacological tool several days after the pairings. This would permit actions of the drug or tool to be dissipated and CNS functions to be restored to normal. Therefore, we employed these procedures in a conditioned reinforcement approach to studying intravenous doses of morphine and amphetamine for their reinforcing capacities.
Conditioned Reinforcement after Noncontingent
Intravenous Doses of Morphine and Amphetamine
At the outset we established conditions under which drug-naive rats would show apparent evidence of primary reinforcement (i.e., would develop self-administration behavior for intravenous doses of either morphine sulfate or dexamphetamine sulfate solutions; Davis & Smith, 1972). Then, using the drug doses found effective in that situation, we demonstrated conditioned reinforcement produced concurrently with acquisition of self-administration behavior for each of these agents by drug-naive rats (Smith & Davis, 1973a; Davis & Smith, 1974a). Subsequently, we tested for the establishment of conditioned reinforcement with a buzzer stimulus when pairing of the buzzer sound was contingent with programmed drug infusion (i.e., drug effects were noncontingent with behavior; Davis et al., 1972; Davis & Smith, 1974a). We will focus our attention herein on this latter paradigm. An initial experiment aimed at validating the conditioned reinforcement approach was to demonstrate a magnitude of reinforcement or dose response relationship for intravenous doses of morphine employed to establish a conditioned reinforcer (Crowder, Smith, Davis, Noel, & Coussens, 1972). The subjects were allowed a baseline period for determining the operant level of bar-pressing. Here there were no drug contingencies associated with such behavior. However, a 0.2-second neutral buzzer stimulus occurred with each response on the lever, concurrently with an injection of saline solution. At the end of the operant-level period, the lever was removed from the chamber. After one hour a second session began with 100 noncontingent morphine injections given concurrently with buzzer presentations. Morphine doses (as the sulfate) were given to three groups at either 0.0032, 0.032, or 0.32 mg/kg in the same volume and duration as for the prior saline doses. These small morphine doses, shown to generate high responding for self-administration (Weeks & Collins, 1971), were delivered randomly without regard to behavior. The sessions lasted about 200 minutes.
After the pairing sessions the lever had remained out of the chambers until the following day, when it was restored at the same time of day as the initial operant level session. Conditions of that first session were repeated: A lever-response produced the buzzer stimulus plus a saline infusion. This comprised a test for establishment of secondary reinforcing potency for the buzzer stimulus. Immediately after this test a further session began in which lever-pressing led to delivery of the buzzer stimulus plus the same dose of morphine as had been given on the previous day. This period was used to detect any subjects that did not respond for morphine solution as a primary reinforcer. Rats not sensitive to morphine as a primary reinforcer could not be expected to develop conditioned reinforcement. The data for six subjects were eliminated on this basis.
Results of the test for conditioned reinforcement are shown in Figure 1. The baseline operant responses of the groups did not differ, but after the buzzer-morphine pairings their responses were elevated in a dose-related manner (ANOVA, p < 0.01); there was a significant linear trend with log dose (p < 0.01). These results confirmed that repeated small intravenous doses of morphine sulfate paired with a buzzer stimulus had imparted reinforcing properties to the buzzer stimulus and that the magnitude of this reinforcement increased linearly with the unit dose of morphine injected. Absence of withdrawal signs indicated that the amount of morphine administered was insufficient to induce acute physical dependence (Coussens, Crowder, & Smith, 1973).
Another validating study (Davis & Smith, 1974a) used a two-lever
condition like the procedure employed by Stein (1958). This was
with dexamphetamine sulfate as the reinforcer at a dose of 0.015 mg/kg;
100 pairing trials were given on Day 2, and the test for conditioned
was on Day 6 rather than Day 3. Half of the subjects received the
stimulus plus saline infusions for responses on the left lever, and the
other half received the same combination for responses on the right
Responding on the inactive lever had no scheduled consequence, but it
recorded as a measure of nonspecific activity. A much higher number of
responses occurred on the reinforced lever than on the other (mean
vs. 17.2), attesting to the specificity of this behavior as measuring
reinforcement. This demonstration eliminated conditioned activation or
disinhibition of motor performance as explanations of the conditioned
More extensive data for dexamphetamine were reported later (Davis,
& Khalsa, 1975) along with similar results for morphine in the
design (see Figure 2).
|Figure 1: Mean number of lever presses during a 6-hour test for conditioned reinforcement with a buzzer-saline infusion contingency 24 hours after 100 buzzer-morphine infusion pairings. Morphine sulfate in the indicated doses was rapidly injected intravenously in a small volume coincident with a 0.5-second buzzer presentation. The lever was out of the chamber during pairings. (n = 7 or 8/group.) Reprinted with permission from Crowder, Smith, Davis, Noel, and Coussens, 1972. Copyright 1972 by Denison University.|
Another study confirmed that the level of responding in a test for conditioned reinforcement did not diminish if the test was delayed further—even to 8 or 16 days after the pairing trials—provided that no extinction procedure intervened (Davis & Smith, 1974a). This finding is in accord with what is known for traditional, nondrug reinforcers used as the basis for secondary reinforcement.
CONDITIONED REINFORCEMENT METHOD IN STUDIES
FOR ANTAGONISM OF MORPHINE OR AMPHETAMINE ACTIONS
The successful application of a conditioned reinforcement method to
measure a dose-related variation in magnitude of reinforcement from
morphine, and the several other initial validation studies, encouraged
our further use of this approach to test whether depletion of CNS
(CA) might prevent the primary reinforcing action of morphine and
Despite indications that this might be the case based on our
studies (Davis & Smith, 1972), we felt the need to verify such data
by means of the conditioned reinforcement paradigm to preclude motor
or performance factors having possibly contaminated the former data.
general procedure of the initial studies described above was adopted
a 4-step, 7-day format; Day 1 consisted of a 6-hour operant baseline
in which lever presses caused saline infusions. On Day 2 the response
were removed from the chambers, and the rats received three
injections of either alpha-methyl-para-tyrosine (AMT) or saline
at 8, 4, and 0 hours before starting an experimental session consisting
of 100 noncontingent pairings of the buzzer stimulus and a rapid
injection of morphine sulfate solution (0.032 mg/kg) on a
programmed schedule. On Day 6 the rats were again placed in the chamber
for the test of conditioned reinforcement under the buzzer-saline
as on Day 1. The interval between second and third stages (Day 2 to Day
6) of the experiment permitted recovery from the depression of brain CA
levels produced by AMT (Davis & Smith, 1974b). On Day 7 the rats
given a test for possible failure to show primary reinforcement with
dose of morphine used. Only one of 20 rats failed this test and was
excluded. The results of the Day 6 test showed that the group of rats
= 9) receiving AMT before the morphine-buzzer pairings did not respond
above their initial operant level on the saline-buzzer condition.
subjects receiving only saline pretreatment prior to morphine-buzzer
showed lever-pressing nearly three times their baseline level, a highly
significant difference (p < 0.001) both from their own baseline and
from the AMT group. These data indicate that AMT blocked the primary
action of morphine and thus prevented the establishment of conditioned
reinforcement associated with the buzzer stimulus, which was clearly
for the saline pretreated rats. As the test for reinforcement was 4
after the dosing with AMT, it is not reasonable to suppose that the
to show reinforcement was caused by an impairment of performance. Thus,
these data added strong confirmation to the self-administration data,
the conclusion that AMT did indeed block the rewarding effects
with small intravenous doses of morphine (Davis & Smith, 1973a).
|Figure 2: Two-lever validation data for self-administration (SA) and conditioned reinforcement (CR) paradigms with rats receiving morphine sulfate in intravenous doses of 0.032 mg/kg during SA or in 100 noncontingent pairings with buzzer 1 day before the CR test. R indicates responses on the reinforced lever for the SA test or the lever producing the buzzer in the CR test; A indicates responses on the lever having no contingencies associated, providing only an index of general activity. Adapted with permission from Davis, Smith, and Khalsa, 1975.|
Parallel studies performed with dexamphetamine sulfate and AMT (Davis & Smith, 1973b), as with morphine, gave results also parallel to those for morphine. Not only the data for self-administration but also those for the conditioned reinforcement procedure showed a blocking of dexamphetamine-associated primary reinforcement.
Further application of the conditioned reinforcement paradigm was directed to the possible discrimination between noradrenergic and dopaminergic components in the effects of AMT on reward from both morphine and dexamphetamine. Both sodium diethyldithiocarbamate (DDC) and U-14,624 (inhibitors of dopamine b-hydroxylase, which is the enzyme that controls the biosynthetic step between dopamine and norepinephrine) were employed for this purpose (see Figure 3). On the basis of studies with DDC and U-14,624 analogous to those with AMT, it appeared that production of a noradrenergic deficit in this manner could prevent the rewarding action of either morphine or dexamphetamine (Davis et al., 1975).
A particular advantage of the conditioned reinforcement test may be seen in the results of an interaction study of morphine and the dopamine-receptor blocking agent haloperidol (Smith & Davis, 1973b). In this case pretreatment with two large doses (5 and 10 mg/kg) of haloperidol prevented reacquisition of morphine self-administration behavior. However, the same and even higher doses of haloperidol failed to impair the responding in the conditioned reinforcement test of rats that had received haloperidol before the morphine-buzzer pairings. Thus, it was concluded that the former results had reflected a motor inhibitory action of haloperidol rather than a blocking of morphine's rewarding action. This is particularly noteworthy in light of the fact that a lower haloperidol dose (0.5 mg/kg) in both paradigms showed evidence for enhancement of the reward potency of morphine.
In contrast to morphine, dexamphetamine-associated reinforcement was
effectively blocked in both experimental designs by haloperidol (Davis
& Smith, 1975). Moreover dopamine-receptor antagonism by means of
also appeared to explain the antagonism of reinforcement associated
two dopaminergic agonists, apomorphine and piribedil (ET-495), both in
self-administration and conditioned reinforcement tests (Davis &
1977). Depletion of brain norepinephrine via U-14,624 treatment,
was ineffective in altering the rewarding potency of apomorphine (see
4). Similarly, a noradrenergic agent, clonidine, which supported both
and conditioned reinforcement, could be blocked from exerting its
reinforcing action by the noradrenergic receptor blocker
A lack of cross-effectiveness both of phenoxybenzamine versus
and piribedil, and of haloperidol versus clonidine, was demonstrated
on the basis of the conditioned reinforcement procedure. These results
constituted a pharmacological validation for the earlier studies of
paper (Davis & Smith, 1977).
|Figure 3: Effects of two dopamine b-hydroxylase inhibitors (DDC and U-14,624) on establishment of a conditioned reinforcer in rats based on intravenous doses (0.032 mg/kg) of morphine sulfate paired 100 times with a neutral buzzer stimulus. Groups C received the drug vehicle treatment, while Groups E received one of the inhibitors before the pairings session. The CR test was 4 days after the pairings. Data are means (± SEM) for groups of 8 subjects. Adapted with permission from Davis, Smith, and Khalsa, 1975.|
Conditioned Reinforcement Method in
Research on Intragastric Ethanol
Ethanol is a considerably weaker reinforcer than is morphine or
even if they are compared by intravenous injection (Smith & Davis,
1974; Smith, Werner, & Davis, 1975a). Moreover, to model its human
use/misuse, the oral rather than intravenous route of administration is
much to be preferred (see also Amit, Smith, & Sutherland, this
In order to avoid problems associated with voluntary oral intake of
solutions by rats while maintaining the intragastric mode of
we developed an intragastric (i.g.) cannulation technique that proved
to self-administration experiments (Smith, Werner, & Davis, 1975b).
The delay imposed by the intragastric route caused a different
relationship for self-administration than resulted if the same doses
made available intravenously (Smith, Werner, & Davis, 1976).
these facts the buzzer stimulus that overlaid self-administered 100
intragastric doses of ethanol was found to act as a conditioned
increasing responding during extinction on saline contingency trials
Werner, & Davis, 1977; see Figure 5). Furthermore, the
pairing of a buzzer stimulus with experimenter-delivered doses of
(25, 50, or 100 mg/kg, i.g.) 50 times per day for 4 days resulted in a
dose-related gradation of responding in a subsequent 10-hour test of
reinforcement on buzzer-saline contingency (see Figure 6).
|Figure 4: Conditioned reinforcement in rats based on intravenous doses of apomorphine (0.06 mg/kg) paired with a neutral buzzer stimulus in a 100-trial session. Group indicated HAL received 5 mg/kg of haloperidol before beginning the pairings session, while group marked U-14 received U-14,624 before pairings. Reprinted with permission from Davis and Smith, 1977. Copyright 1977 by Pergamon Press Inc.|
Experiments regarding the role of brain catecholaminergic systems in
the rewarding action of ethanol were conducted in analogous fashion to
those described above for morphine and dexamphetamine. The differences
were the use of the intragastric route and a longer (10-hour) session
acquisition of self-administration behavior and in testing for
reinforcement. First studied were AMT and U-14,624, which blocked
self-administration (Davis, Smith, & Werner, 1978). This was
by a similar study of FLA-57, another inhibitor of dopamine b-hydroxylase,
which was fully effective against both reacquisition of
behavior and responding in the test for conditioned reinforcement
Werner, & Smith, 1979).
|Figure 5: Demonstration of effects on extinction-responding of a conditioned reinforcer established by intragastric doses of 25 or 100 mg/kg of ethanol given contiguously with a buzzer stimulus during five daily 10-hour self-administration sessions. Data are means (± SEM) of lever responses under extinction on Day 6. Buzzer and saline injection were contingent upon lever pressing in one-half of each dose group, while the other half received saline injection but no buzzer. Reprinted with permission from Smith, Werner, and Davis, 1977. Copyright 1977 by Springer-Verlag.|
Research on conditioned reinforcement and opiates prior to or at the time of our initial studies (Schuster, & Woods, 1968; Stolerman & Kumar, 1972; Wikler, Pescor, Miller, & Norrell, 1972) employed physically dependent organisms that required a lengthy dependence induction phase. Thus, our data first demonstrated conditioned reinforcement in nondependent subjects. Prior researchers also had required an operant-response-contingent relationship between the neutral stimulus (to become the conditioned reinforcer) and the primary reinforcer, morphine. We introduced the classically conditioned relationship. A reinforcement interpretation requires evidence of learning. The conditioned reinforcer must be validly demonstrated to control the dependent variable--in this case the lever-response or drinking rate—as a learned phenomenon. Since the putative conditioned reinforcers in the early studies cited were tested during opiate abstinence, increased responding could have been caused by disinhibitory factors (such as increased motor or drinking activity) related to withdrawal stress or facilitation of locomotor activity or drinking patterns introducing performance artifacts. Therefore, at the time of our studies, no unconfounded data were available which clearly showed that truly learned, secondary reinforcement phenomena were demonstrable in drug-reward research.
Thus, our first efforts were to demonstrate opiate-based primary
and conditioned reinforcement in nondependent organisms in such ways as
to avoid both dependency and performance factors probably contaminating
previous data. Next, we undertook to develop valid measures for the
of drug-based conditioned reinforcement, valid being defined as
methods uncontaminated by performance variables. This was accomplished
by our Pavlovian pairing method which incorporated a new
instrumental response as its basic measure. Furthermore, we examined
reliability of our results in terms of ability to replicate our data.
was accomplished not only by direct replications of our original
but also by means of cross-reliability studies—by showing magnitude
of reinforcement effects, by delayed testing for acquisition of
reinforcer, and by systematic replication showing conditioned
associated with orthogonal primary reinforcers (i.e., other drugs such
as amphetamine and apomorphine and later ethanol).
|Figure 6: Dose-relationship in a test for conditioned reinforcement with groups that received 0, 25, 50, or 100 mg/kg intragastric doses of ethanol paired with a buzzer stimulus. Pairings were 50 per day for 4 days with the lever removed from the apparatus. The conditioned reinforcement test was a 10-hour session with levers restored and a buzzer-saline contingency like the operant baseline period. Reprinted with permission from Smith, Werner, and Davis, 1977. Copyright 1977 by Springer-Verlag.|
Beyond these methodological considerations there is a considerable significance of drug-based conditioned reinforcement to the clinical bases of drug abuse syndromes. We have described animal experiments directed to the analysis of roles played by conditioned reinforcers in maintenance and/or relapse to drug-seeking behavior (Davis & Smith, 1976). Not only the latter research area but also the utilitarian applications of conditioned reinforcement as a measure of reward from abused drugs merit further attention and exploitation by researchers on drug abuse.
This research was supported by research grants MH 13570, MH 11295, and DA 00018 from the National Institute of Mental Health/National Institute on Drug Abuse and by the Research Institute of Pharmaceutical Sciences of the University of Mississippi.
Coussens, W. R., Crowder, W. F., & Smith, S. G. (1973). Acute physical dependence upon morphine in rats. Behavioral Biology, 8, 533-543.
Crowder, W. F., Smith, S. G., Davis, W. M., Noel, J. T., & Coussens, W. R. (1972). Effect of morphine dose size on the conditioned reinforcing potency of stimuli paired with morphine. Psychological Record, 22, 441-448.
Davis, W. M., & Smith, S. G. (1972). Alpha-methyltyrosine to prevent self-administration of morphine and amphetamine. Current Therapeutic Research, 14, 814-819.
Davis, W. M., & Smith, S. G. (1973a). Blocking of morphine based reinforcement by alpha-methyltyrosine. Life Sciences, 12, 185-191.
Davis, W. M., & Smith, S. G. (1973b). Blocking effect of alpha-methyltyrosine on amphetamine based reinforcement. Journal of Pharmacy and Pharmacology, 25, 174-177.
Davis, W. M., & Smith, S. G. (1974a). Behavioral control exerted by an amphetamine based conditioned reinforcer. In J. M. Singh & H. Lal (Eds.), Drug addiction Volume 3: Neurobiology and influences on behaviors (pp. 209-219). New York: Stratton Intercontinental Medical Book Corp.
Davis, W. M. , & Smith, S. G. (1974b). Noradrenergic basis for reinforcement associated with morphine action in nondependent rats. In J. M. Singh & H. Lal (Eds.), Drug addiction Volume 3: Neurobiology and influences on behavior (pp. 155-168). New York: Stratton Intercontinental Medical Book Corp.
Davis, W. M., & Smith, S. G. (1975). Effect of haloperidol on (+)-amphetamine self-administration. Journal of Pharmacy and Pharmacology, 27, 540-542.
Davis, W. M. , & Smith, S. G. (1976). Role of conditioned reinforcers in the initiation, maintenance and extinction of drug-seeking behavior. Pavlovian Journal of Biological Science, 11, 222-236.
Davis, W. M., & Smith, S. G. (1977). Catecholaminergic mechanisms of reinforcement: Direct assessment by drug self-administration. Life Sciences, 20, 483-492.
Davis, W. M., Smith, S. G., & Crowder, W. F. (1972). Morphine-based conditioned reinforcement. Abstracts of Volunteer Papers, 5th International Congress on Pharmacology, p. 52.
Davis, W. M., Smith, S. G., & Khalsa, J. H. (1975). Noradrenergic role in the self-administration of morphine or amphetamine. Pharmacology Biochemistry & Behavior, 3, 477-484.
Davis, W. M., Smith, S. G., & Werner, T. E. (1978). Noradrenergic role in the self-administration of ethanol. Pharmacology Biochemistry & Behavior, 9, 369-374.
Davis, W. M., Werner, T. E., & Smith, S. G. (1979). Reinforcement with intragastric infusions of ethanol: Blocking effect on FLA 57. Pharmacology Biochemistry & Behavior, 11, 545-548.
Kumar, R. (1972). Morphine dependence in rats: Secondary reinforcement from environmental stimuli. Psychopharmacologia, 25, 332-338.
Liebman, J. M. (1983). Discriminating between reward and performance: A critical review of intracranial stimulation methodology. Neuroscience & Biobehavioral Reviews, 7, 45-72.
Schuster, C. R., & Woods, J. H. (1968). The conditioned reinforcing effects of stimuli associated with morphine reinforcement. International Journal of the Addictions, 3, 223-230.
Smith, S. G., & Davis, W. M. (1973a). Behavioral control by stimuli associated with acquisition of morphine self-administration. Behavioral Biology, 9, 777-780.
Smith, S. G., & Davis, W. M. (1973b). Haloperidol effects on morphine self-administration: Testing for pharmacological modification of the primary reinforcement mechanism. Psychological Record, 23, 215-221.
Smith, S. G., & Davis, W. M. (1974). Intravenous alcohol self-administration in the rat. Pharmacological Research Communications, 6, 397-402.
Smith, S. G., Werner, T. E., & Davis, W. M. (1975a). Intravenous drug administration in rats: Substitution of ethyl alcohol for morphine. Psychological Record, 25, 17-20.
Smith, S. G., Werner, T. E., & Davis, W. M. (1975b). Technique for intragastric delivery of solutions: Application for self-administration of morphine and alcohol by rats. Physiological Psychology, 3, 220-224.
Smith, S. G., Werner, T. E., & Davis, W. M. (1976). Comparison between intravenous and intragastric alcohol self-administration. Physiological Psychology, 4, 91-93.
Smith, S. G., Werner, T. E., & Davis, W. M. (1977). Alcohol-associated conditioned reinforcement. Psychopharmacology, 53, 223-226.
Stein, L. (1958). Secondary reinforcement established with subcortical stimulation. Science, 127, 466-467.
Stolerman, I. P., & Kumar, R. (1972). Secondary reinforcement in opioid dependence. In J. M. Singh, L. H. Miller, & H. Lal (Eds.), Drug addiction, Volume 1: Experimental pharmacology (pp. 49-50). New York: Futura.
Valenstein, E. S. (1964). Problems of measurement and interpretation with reinforcing brain stimulation. Psychological Reviews, 71, 415-437.
Weeks, J. R., & Collins, R. J. (1971). Primary addiction to morphine in rats. Federation Proceedings, 30, 277.
Wikler, A., Pescor, F. T., Miller, D., & Norrell, H. (1971). Persistent potency of a secondary (conditioned) reinforcer following withdrawal of morphine from physically dependent rats. Psychopharmacologia, 20, 103-117.
Click here to enter the Addiction Science Network Discussion Forum
©1999-2009 Addiction Science Network
This page was last revised
06 April 2009 15:31 EDT.
Send comments to: feedback@AddictionScience.net
Report technical problems