Higher order conditioning of a taste aversion

Two groups of rats each received 5 drops of NaCl solution from a dropper placed directly inside the mouth. The experimental group was then injected with lithium chloride to establish a conditioned taste aversion. The control group was injected 24 h later. After a recovery day the above procedures were repeated. On the next day both groups received 5 drops of a saccharin solution followed immediately by 5 drops of the NaCl solution. Subsequent preference tests established that the experimental group had learned an aversion to the saccharin solution as a result of its pairing with the NaCl solution which had previously been associated with poisoning. These results demonstrate that higher order conditioning of a taste aversion can be established using tastes as both the first-order stimulus and the second-order stimulus.     

Stimulus selection in discriminative taste-aversion learning in the rat

In three experiments, rats were presented compound solutions consisting of a common element, saccharin, mixed with one of two different flavor elements, cinnamon and wintergreen. Rats in the experimental groups consistently received a toxicosis-inducing injection following one compound solution but not following the other compound solution. Rats in the control groups received toxicosis-inducing injections half the time following each of the compound solutions. After training in each experiment, there were tests for conditioning to the saccharin alone. The experimental groups drank significantly more than the control groups, indicating that the aversion to the partially reinforced saccharin in isolation was less when the different flavor cues were more highly correlated with reinforcement. In Experiment III, there was also a test for conditioning to the cinnamon or wintergreen flavor alone. The experimental group drank significantly less of the continuously reinforced flavor than the control group did of the partially reinforced flavor. These results are similar to those reported within more traditional conditioning paradigms.     

Taste aversion learning and schedule-induced polydipsia in rats

Experiment 1 sought to determine whether schedule-induced drinking could be abolished by means of a taste aversion. Polydipsic rats were given access to a .4% saccharin solution while they were exposed to an intermittent food schedule. Immediately after the session, they received an intraperitoneal injection of either lithium chloride or sodium chloride. Following a recovery day with water in the experimental chamber, the animals were again exposed to the saccharin solution. The poisoned animals (lithium chloride) drank very little saccharin compared to the control animals (sodium chloride), indicating that they had learned a taste aversion in only one conditioning trial. Experiment 2 established that polydipsic rats can learn a taste aversion despite a long delay between schedule-induced saccharin consumption and poisoning, and that the delay gradient displayed by polydipsic rats is similar to that observed in thirst-motivated rats.     

Effects of expected vs. unexpected proximal US preexposure on taste-aversion learning

Animals were first conditioned to expect lithium treatment following exposure to one taste solution (the CS+) and to expect no drug treatment following exposure to another flavor (the CS?). All subjects then received a saccharin taste-aversion conditioning trial. In Experiment 1, this conditioning trial was preceded 0, 1, 2, 4, or 6 h earlier by exposure to the CS+ flavor for independent groups. The CS+ exposure attenuated saccharin aversion learning if it occurred immediately before the saccharin conditioning trial but not if it occurred 1 h or more before conditioning. In Experiment 2, the saccharin conditioning trial was preceded 3 or 4.5 h earlier by a lithium injection. This proximal US preexposure injection was either unannounced (Li) or preceded by exposure to the CS+ (CS+Li) or the CS? (CS?Li) stimuli. The US preexposure attenuated saccharin aversion learning in all cases. However, the interference effect was less when the preexposure injection was expected (CS+Li) than when it was unexpected (CS?Li). This outcome could not be explained in terms of direct effects of the CS+ and CS? stimuli on the saccharin conditioning trial, and shows that the proximal US preexposure effect is a function of not only the drug dosage and preexposure interval, but also the anticipation of the drug pretreatment.     

Amount of solution drunk is a factor in the establishment of taste aversion

Experiment I demonstrated that the strength of a rat’s aversion to saccharin is a direct function of the amount of saccharin it consumed prior to poisoning. Using Kalat and Rozin’s (1973) procedure, Experiment II showed that results consistent with a “learned-safety” theory of taste aversion appear to depend on whether rats drink most saccharin on their first or second exposure to the solution prior to poisoning. Experiment III demonstrated that when animals drank equal amounts of saccharin solution on each of two exposures prior to poisoning, evidence strongly confirming the “learned-safety” theory was obtained. These experiments together demonstrate the importance of amount of solution drunk in the determination of taste aversion.     

Effects of proximal unconditioned stimulus preexposure on ingestional aversions learned as a result of taste presentation following drug treatment

Taste aversions were conditioned by exposing subjects to a 1.0% saccharin solution 30 min after an injection of lithium chloride. The aversion learning was disrupted if subjects had also received an additional lithium injection some time earlier (Experiments 1–3). This interference effect of US preexposure was a decreasing function of the preexposure interval, beyond the optimal interval (105 min) for observing the phenomenon (Experiment 1), and was directly related to the dose of the preexposure injection (Experiment 2). No interference with conditioning occurred at short (e.g., 30-min) preexposure intervals (Experiment 1), probably because under these circumstances the preexposure injection itself conditioned a strong aversion (Experiment 4). At moderate (105-min) but not at short (30-min) preexposure intervals, the interference with aversions learned as a result of taste exposure following drug injection was comparable to the interference with learning in a more conventional forward conditioning procedure (Experiments 3 and 4). These findings are similar to previously documented effects of proximal CS- and US-preexposure and are consistent with recent stimulus rehearsal and opponent-process theories.     

Injected flavor as a CS in the conditioned aversion preparation

Three experiments were conducted to determine the effectiveness of intravenous (IV) flavor injections in the formation of conditioned taste aversions and in the attenuation of neophobia. In Experiment 1, two groups of rats were permitted to drink either a .1% saccharin solution or tap water followed immediately by IV injections of lithium chloride (LiCl), and two more groups were given IV injections of a 2% saccharin solution followed immediately by IV injections of either LiCl or distilled water. Injected flavor did not serve as an effective CS for the conditioning of an aversion to .1% saccharin. The second experiment employed a two-bottle procedure to detect attenuation of neophobia using the injected-flavor technique. It was found that, whether saccharin had been injected intravenously (2%), injected intraperitoneally (2% IP), or orally consumed (.1%), neophobia for .5% saccharin was attenuated equally relative to controls. CS-US intervals were manipulated in the final experiment such that IP injections of 2% saccharin solution were followed 0–480 min later by IP injections of LiCl. In this case, it was shown that injected flavor (2% saccharin) could act as an effective CS if the US was delayed (optimally about 120 min) and when the test solution was .1% saccharin. The delay gradient found in Experiment 3 was interpreted as a generalization gradient where optimum conditioning was displayed at the point where the concentration of saccharin circulating in the animal at the time of illness onset most closely matched the concentration of the test solution.     

Specificity of cue to consequence in aversion learning in the rat: Control for US-induced differential orientations

In four experiments, each using a single conditioning trial, rats avoided a light more than a saccharin solution after these stimuli had been paired with footshock, whereas saccharin was avoided more than the light after these stimuli had been paired with lithium injection. The use of a single conditioning trial precludes possible US-induced differential orientations from influencing which stimuli will be associated on the conditioning trial. This cue-consequence specificity effect was obtained even when subjects conditioned with lithium received a non-contingent footshock prior to the test session, and when subjects conditioned with footshock received a noncontingent lithium injection before testing (Experiments 2–4). Weak aversions to the light in rats given a light-lithium pairing and noncontingent footshock and to the saccharin in subjects that received a saccharin-footshock pairing and noncontingent lithium administration were obtained in Experiment 2. However, these weak aversions were not obtained when subjects were given three nonreinforced exposures to the test chamber before the test session (Experiments 3 and 4). These results indicate that US-induced differential orientations do not mediate the cue-consequence effect in aversion learning.     

Saccharin aversions induced by lithium chloride toxicosis in a backward conditioning paradigm

Seven experimental groups of seven rats each were allowed to consume saccharin solution at different times relative to intubation of lithium chloride solution. Six backward conditioning (BWD) groups were intubed 0.5, 1, 2, 3, 4, and 8 h before saccharin consumption, and a forward conditioning (FWD) group was intubed 0.5 h after saccharin consumption. A no-lithium control group of 14 rats received no intubation. Only the 0.5-h FWD and the 0.5-h BWD groups showed an aversion to saccharin relative to the no-lithium controls. The aversion to saccharin in the 0.5-h FWD group was more pronounced than that in the 0.5-h BWD group. This shows that the aversive effects of lithium toxicosis dissipate far sooner than the aversive effects of X-irradiation.     

Taste quality and extinction of a conditioned taste aversion in rats

Rats (Rattus norvegicus) that received a taste cue (saccharin, saline, quinine, or sucrose) paired with a lithium chloride (LiCl) injection displayed a robust decrease in consumption of that taste, relative to controls that had the taste unpaired with LiCl. Consumption of the paired taste increased with each nonreinforced presentation (i.e., extinction). After asymptotic extinction, rats that had had a 0.1% saccharin cue paired with LiCl consumed less of the saccharin solution than did controls. A similar data pattern was observed with a 10% sucrose solution. These results are consistent with the view that some aspect of the excitatory CS-US association remains after extinction. On the other hand, rats that had a bitter (0.005% or 0.001% quinine) or salty (1% or 0.5% saline) solution paired with LiCl drank similar amounts of the fluid as controls after asymptotic extinction treatment. Together, these experiments suggest that a taste that is either sweet or preferred is required in order to demonstrate the chronic decrease in fluid consumption after extinction treatment. The data suggest that the conditioning experience prevents the later development of a preference for the sweet taste, rather than there being a retained aversion that suppresses fluid consumption.     

The effect of drug habituation before and after taste aversion learning in rats

In Experiment I, rats received eight habituation injections of either lithium chloride (LiCl) or sodium chloride (NaCl), then two aversion training trials in which access to saccharin solution was followed by LiCl injections, and finally eight extinction trials with saccharin but no injections. The rats habituated to LiCl showed less aversion to saccharin during training and extinction. In Experiment II, rats received two aversion training trials, then eight habituation trials to either LiCl or NaCl, then eight extinction trials, four more aversion training trials, and eight more extinction trials. The rats habituated to LiCl did not differ during the first extinction period from those habituated to NaCl, but showed less aversion to saccharin during the second training and extinction periods. Consequently, habituation to LiCl reduces the learning of an aversion to saccharin but does not reduce the performance of a previously learned aversion.     

Anticipatory contrast: Within-subjects analysis

Contrast in consummatory behavior occurs readily when a less preferred substance follows a preferred substance. A previous experiment indicated that contrast in consummatory behavior may also develop when a less preferred substance precedes a preferred substance in brief daily exposures. In the present experiment, the same animals sometimes received 0.15% saccharin followed, 15 sec later, by 32% sucrose (.15–32) and sometimes received 0.15% saccharin followed by the same 0.15% saccharin solution (.15-.15). One solution pair was given each day, and the two conditions, .15–32 and .15-.15, occurred in alternation across days. The two different solution conditions were correlated with different cues. Saccharin intake from the first tube was lower when the second tube contained 32% sucrose than when it contained .15% saccharin, both in original discrimination training and following a reversal of the cue-solution pairings. These results support the conclusion that contrast in this situation is based on the anticipation of the impending 32% sucrose each day, rather than on a retrograde comparison with the 32% sucrose received the previous day. These data are considered in terms of Pavlovian conditioning outcomes when the CS is a stimulus with hedonic value.     

Licking one saccharin solution for access to another in rats: Contingent and noncontingent effects in instrumental performance

Separation of the contingent and noncontingent effects of a schedule on amount of instrumental responding is desirable but difficult in schedules that involve instrumental and contingent responses that are either highly probable or very similar. Three studies in which rats were required to lick a solution of .1% saccharin for access to a preferred solution of .4% saccharin showed that neither single nor paired operant baselines of the instrumental response allowed accurate separation of the contingent and noncontingent effects of a fixed-ratio schedule. Two within-subject yoking procedures provided the best baselines of noncontingent effects: the massed baseline measured amount of .1% licking when each subject received free access to the total amount of .4% licking it obtained at asymptote under the schedule; the matched baseline measured .1% licking when each subject received the same access to the .4% solution, but presented in the intermittent pattern obtained during the schedule. Of the three algebraic models used to predict noncontingent effects, the substitution model was most promising, but still not adequate. The procedure of a between-subjects yoked control was also not effective.     

The effect of nonreinforced stimulus exposure on the strength of a conditioned taste aversion as a function of retention interval: Do latent inhibition and extinction involve a shared process?

Two experiments examined the effects of nonreinforced flavor exposure on the strength of a conditioned taste aversion. Rats were conditioned by pairing maple flavor with LiC1. Prior to or subsequent to this pairing, some animals received nonreinforced exposure to either maple or saccharin. In separate subjects, preference for maple was tested 1 or 21 days after the last training episode. In the first experiment, the nonreinforced stimulus exposure occurred before conditioning (latent inhibition, or LI, procedure); in the second experiment, the nonreinforced exposure occurred after conditioning (extinction, or EXT, training). In both experiments, nonreinforced exposure to maple or saccharin reduced the magnitude of a conditioned maple aversion when testing occurred soon after conditioning. When testing was delayed, however, the attenuation due to nonreinforced saccharin exposure dissipated, both with the LI procedure and with EXT. In contrast, the nonreinforced exposure to maple was found to attenuate conditioned reactions at both short and long retention intervals. The release from generalized LI and spontaneous recovery following generalized EXT training are discussed in terms of retrieval processing. The possibility that the same mechanism may underlie LI and EXT is considered.     

Impact of brief or extended extinction of a taste aversion on inhibitory associations: Evidence from summation,retardation, and preference tests

In five conditioned taste aversion experiments with rats, summation, retardation, and preference tests were used to assess the effects of extinguishing a conditioned saccharin aversion for three or nine trials. In Experiment 1, a summation test showed that saccharin aversion extinguished over nine trials reduced the aversion to a merely conditioned flavor (vinegar), whereas three saccharin extinction trials did not subsequently influence the vinegar aversion. Experiment 2 clarified that result, with unpaired controls equated on flavor exposure prior to testing; the results with those controls suggested that the flavor extinguished for nine trials produced generalization decrement during testing. In Experiment 3, the saccharin aversion reconditioned slowly after nine extinction trials, but not after three. Those results suggested the development of latent inhibition after more than three extinction trials. Preference tests comparing saccharin consumption with a concurrently available fluid (water in Experiment 4, saline in Experiment 5) showed that the preference for saccharin was greater after nine extinction trials than after three. However, saccharin preference after nine extinction trials was not greater, as compared with that for either latent inhibition controls (Experiments 4 and 5) or a control given equated exposures to saccharin and trained to drink saline at a high rate prior to testing (Experiment 5). Concerns about whether conditioned inhibition has been demonstrated in any flavor aversion procedure are discussed. Our findings help explain both successes and failures in demonstrating postextinction conditioned response recovery effects reported in the conditioned taste aversion literature, and they can be explained using a memory interference account.     

Poison avoidance and patch (location) selection in rats

Thirsty rats were tested on a four-armed radial maze with three water locations and one distinctive taste location (saccharin). Rats that were injected with lithium chloride after drinking a novel saccharin solution visited the saccharin location less than did unpoisoned animals, primarily during the later portions of the test sessions. When saccharin was moved to a different location, previously poisoned rats rapidly avoided the new saccharin location and increased visits to the original saccharin location, now rebaited with water. A similar pattern of learned avoidance and approach was obtained in Experiment 2 with three water locations and one vacant location (no water). These results indicate that: (1) sampling the contents of alternative patches mediates both learning to avoid the location of an aversive substance and returning to a newly viable patch, and (2) avoiding the location of a novel substance after a single poisoning occurs because the location does not contain an edible substance, not because of an aversion conditioned to environmental cues.     

The role of amount consumed in flavor preexposure effects and neophobia

Rats received either a small or a large amount of a novel saccharin solution prior to a conditioning episode in which the saccharin flavor was paired with lithium-induced toxicosis. When the time between the preexposure and the conditioning episode was 3.5 h, both the small and large amounts of saccharin decremented conditioning. However, when the time between the preexposure and the conditioning episode was 23.6 h, only the consumption of a large amount of saccharin decremented conditioning. In a second experiment, rats received either a short or a long exposure to the novel saccharin solution and were then given a choice test between the saccharin and water. Rats given the choice test immediately following their preexposure to saccharin avoided consuming it. If they received the choice test 4 h later, they consumed more of the saccharin irrespective of the length of the preexposure. In contrast, when they were given a choice test 24 h after the preexposure, only rats that had received a long preexposure displayed a significant preference for the saccharin. The present results demonstrate that the flavor preexposure effect and the attenuation of neophobia are both determined by the time between preexposure to the novel flavor and the conditioning episode or choice test and the amount of preexposure to the novel flavor. These findings are discussed in relation to the information-processing model developed by Wagner (1976, 1978) and are used to provide an account, in terms of this model, of the delay-gradient seen in flavor-aversion learning.     

Second-order conditioning detects unexpressed morphine-induced salt aversion

Morphine failed to condition a salt taste aversion at a dose (15 mg/kg) sufficient to produce a robust aversion to a saccharin taste. Indeed, three different concentrations of salt (1%, 1.5%, and 2%) paired with the same morphine dose yielded no direct evidence for conditioned aversion. Yet, when a novel saccharin taste was paired in compound with the previously conditioned salt conditioned stimulus, we found evidence for a conditioning to the saccharin cue alone in three separate experiments. Control groups eliminated alternative accounts such as neophobia and differential exposure to morphine. Combined, these findings indicate that morphine conditioned a salt aversion. Although this aversion was not directly expressed, a second-order conditioning procedure was able to provide a more sensitive index of conditioning.     

Foraging on the radial-arm maze: Effects of altering the reward at a target location

Thirsty rats were trained to collect small water rewards from the end of each arm of an eight-arm radial maze. During these training trials and subsequent testing trials, the subjects were allowed to choose a maximum of eight arms. “Preference” for a target maze location was studied by noting when, in the sequence of eight choices, the target was selected. During testing, when one maze location was consistently devoid of water, rats decreased their preference for this arm over trials (Experiment 1). Similarly, rats that learned a saccharin-lithium association demonstrated lower preferences for a maze location that consistently held the conditioned saccharin solution. This was true for animals that received saccharin-lithium conditioning on the maze (Experiment 3A) and for animals conditioned to saccharin in a separate context (Experiment 3B). An increase in preference for a target maze location consistently containing a sweet chocolate milk solution was observed in animals that were water- and food-deprived (Experiment 2). These studies demonstrate that animals will modify their responses toward (preferences for) maze locations that predictably contain an altered reward.     

Transfer of a taste aversion from food to water under various states of deprivation

The importance of ingestive contexts (feeding and drinking) and deprivation states to rats’ transfer of a taste aversion were examined, In Experiments 1 and 2, rats were trained with novel saccharin-treated foods while either food deprived or food and water deprived, They were then tested with a 1,0% saccharin solution while either water deprived or food and water deprived, Comparable aversions to the solution were displayed regardless of deprivation states, Two further experiments examined transfer to a .1% saccharin solution in conjunction with deprivation state change, When both stimulus properties and deprivation were widely discrepant from training to test, reduced transfer was noted. The results suggest that stimulus similarity was a stronger controlling variable than deprivation state similarity in facilitating the transfer of an aversion from a feeding context to a drinking context, The results were viewed as being consistent with the known parameters affecting generalization gradients.