Pigeons’ memory for empty time intervals marked by visual or auditory stimuli

In Experiment 1, pigeons were trained to discriminate the duration (2 or 8 sec) of an empty interval separated by two 1325-Hz tone markers by responding to red and green comparison stimuli. During delay testing, a choose-short bias occurred at 1 sec, but a robust choose-long bias occurred at 9 sec. Responding in the absence of tone markers indicated that the pigeons were attending to the markers and not simply timing the total trial duration. The birds were then trained to match short (2-sec) or long (8-sec) empty intervals marked by light to blue/yellow comparisons. For both visual and auditory markers, delay testing produced a choose-short bias at 1 sec and a choose-long bias at 9 sec. In Experiment 2, the pigeons were shifted from a fixed to variable intertrial intervals (ITI) within sessions. On trials with tone markers, the duration of both the empty interval and the preceding ITI affected choice responding. On trials with light markers, only the duration of the empty interval influenced choice responding. Subsequent delay testing in the context of variable ITIs replicated the memory biases previously obtained. In Experiment 3, performance was assessed at various delay intervals on trials in which either the first or the second marker was omitted. The data from these omission tests indicated that the first marker initiated timing but that the second marker sometimes initiated the timing of a new interval. Explanations of these effects in terms of the internal clock model of timing are discussed, and a simple quantitative model of the delay interval data is tested.     

Memory for empty time intervals in pigeons

Pigeons were trained to discriminate short (2 sec) and long (8 sec) empty intervals that began each trial. In group consistent, onset of an empty interval was marked by a brief presentation of red keylight, and termination of the interval was marked by a brief presentation of green keylight. In group inconsistent, red and green served equally often as the first and second markers across trials. Testing revealed that, in group consistent, (1) birds were sensitive to the relation between marker color and marker type and (2) presentation of the second marker did not initiate timing a new interval. Testing also revealed a robust choose-long effect at delays longer than the training delay and indifference between the comparisons on no-sample trials. Both of the latter findings differ from those typically obtained when filled intervals are employed. It was concluded that pigeons process filled and empty intervals differently.     

Memory for duration in pigeons: Dissociation of choose-short and temporal-summation effects

Five groups of pigeons were trained in a symbolic choice-matching feast involving short (2-sec) and long (10-sec) durations of houselight as samples. Four groups also received training with a second set of samples: line orientations or 2- and 10-sec presentations of keylight. The type of sample-to-comparison mapping varied across groups. Although only two of the five groups demonstrated a choose-short effect (a tendency to choose the comparison associated with a short sample at longer delays), all groups demonstrated temporal summation (a tendency to respond on the basis of the combined duration of two successively presented samples). Moreover, the magnitude of temporal summation was equivalent in groups that did and did not-demonstrate a choose-short effect. The results suggest that the processes underlying the perception of sample duration remain invariant across different sample-to-comparison mapping arrangements, but that the memory code used to retain temporal information varies.     

Systematic errors in pigeons’ memory for event duration: Interaction between training and test delay

Pigeons trained to choose different stimuli following short- and long-duration signals make disproportionately more “short” choices (i.e., “choose-short errors”) following an increase in the retention interval and more “choose-long errors” following a decrease in this delay. The present experiment provided a systematic investigation of how these selective errors depend on the relationship between the training delay and the test delay. Pigeons were first trained with a 0-sec delay between the signal (2- or 8-sec food presentations) and the choice stimuli (red- and blue-lit keys). On subsequent test trials with 5- and 10-sec delays, choose-short errors predominated. Next, the birds were trained with a constant 10-sec delay and then tested with shorter or longer delays on some trials. The birds now responded accurately and without selective errors at the 10-sec training delay, but made choose-long errors at shorter delays and choose-short errors at longer delays. Finally, the birds were trained with a constant 20-sec delay and then tested with shorter and longer delays. Choose-long errors again appeared at shorter test delays, choose-short errors at longer test delays, and no differential errors at the 20-sec training delay. The selectivity of these errors generally increased with the absolute difference between the training and test delay. Theoretical implications of these results are discussed.     

Pigeons’ memory for event duration: Intertrial interval and delay effects

The effects of within-session variations in the intertriai interval (ITI) and delay on pigeons’ memory for event duration were studied in delayed symbolic matching-to-sample tasks. Pigeons were trained to peck one color following a long (8 sec) sample and another color following a short (2 sec) sample. In the first three experiments, the baseline conditions included a 10-sec delay (retention interval) and a 45-sec ITI. During testing, the delay was varied from 0 to 20 sec, and the ITI that preceded the trial was varied from 5 to 90 sec. When the ITI and delay were manipulated separately (Experiments 1 and 2), the pigeons displayed a choose-short tendency when the delay was longer than 10 sec or when the ITI was longer than 45 sec, and a choose-long tendency when either the delay or the ITI was shorter than these baseline values. These effects occurred whether the sample was food access or light. When the ITI and delay were manipulated together, the pigeons showed a large choose-long error tendency when the short delay was tested together with a short ITI, and no systematic error tendency when the short delay was tested together with a longer ITI. A very large choose-short error tendency emerged on trials with a long delay and a long ITI; a reduced choose-short tendency was present when the long delay was presented together with a short ITI. In Experiment 4, the baseline conditions were a 0-sec delay and a 45-sec ITI. In this case variations in the ITI had a smaller and unidirectional effect: the pigeons showed a choose-long error tendency when the ITI was decreased, but no effect of ITI increases. Two hypotheses were proposed and discussed: (1) that pigeons judge sample durations relative to a background time composed of the ITI and delay, and (2) that the delay and ITI effects might arise from a combination of subjective shortening and proactive effects of samples from previous trials.     

The role of keypecking during filled intervals on the judgment of time for empty and filled intervals by pigeons

Pigeons were trained in a within-subjects design to discriminate empty intervals (bound by two 1-sec visual markers) and filled intervals (a continuous visual signal). The intervals were signaled by different visual stimuli and they required responses to different sets of comparison stimuli. In Experiment 1, empty intervals were judged longer than filled intervals. The difference between the point of subjective equality (PSE) for the empty intervals and the PSE for filled intervals increased as the magnitude of the anchor-duration pairs increased. Although there was more pecking during filled intervals than during empty intervals, there was no significant correlation between pecking during filled intervals and the value of the PSE. In Experiment 2, empty intervals continued to be judged longer than filled intervals, even when pigeons were required to refrain from pecking during filled intervals. Keypecking per se does not appear to play an important role in the empty-filled timing difference.     

Memory for number of light flashes in the pigeon

Two groups of pigeons were trained to perform symbolic delayed matching-to-sample at a 0-sec delay with sample stimuli that consisted of sequences of light flashes. The sequences varied in number but not time for one group (number group) and in time but not number for the other group (time group). When retention was tested at delays up to 10 sec in Experiment 1, a choose-small effect was found in the number group, and a choose-long effect was found in the time group. Transfer tests between number and time samples in Experiment 2 supported the hypothesis that pigeons were discriminating between the number of light flashes at the end of sample sequences in Experiment 1. It was concluded that pigeons in both the number and the time groups were discriminating between number of flashes and that the apparent choose-long effect was actually a choose-small effect. The implications of these findings for the mode-control model of counting and timing (Meck & Church, 1983) were discussed.     

Event-duration discrimination by pigeons: The choose-short effect may result from retention-test novelty

Pigeons trained on a conditional event-duration discrimination typically “choose short” when retention intervals are inserted between samples and comparisons. In two experiments, we tested the hypothesis that this effect results from ambiguity produced by the similarity of the novel retention intervals and the familiar intertrial interval by training pigeons with retention intervals from the outset and, for one group, in addition, making retention intervals distinctive from the intertrial intervals. In Experiment 1, when the retention intervals (0–4 sec) were not distinctive from the intertrial intervals, the pigeons did not show a clear choose-short effect even when extended retention intervals (8 sec) were introduced. When the retention intervals were distinctive, the pigeons showed a choose-long effect (they appeared to time through the retention interval), but it was relatively weak until the retention intervals were extended to 8 sec. In Experiment 2, when pigeons were discouraged from timing through the retention intervals by making the intertrial intervals and retention intervals salient distinct events and using long (up to 16-sec) retention intervals in training, parallel retention functions were found. It appears that when ambiguity is removed, forgetting by pigeons does not occur by the process of subjective shortening. These experiments suggest that the accurate interpretation of results of animal memory research using differential-duration samples must consider the novelty of the retention intervals on test trials as well as their similarity to other trial events.     

The effect of filled and empty intervals on clock and memory processes in pigeons

According to the mixed memory model (Penney, Gibbon, & Meck, Journal of Experimental Psychology: Human Perception and Performance, 26, 1770–1787, 2000), different clock rates for stimuli with different nontemporal properties must be stored within a single reference memory distribution in order to detect a difference between the clock rates of the different signals. In Experiment 1, pigeons were trained in a between-subjects design to discriminate empty intervals (bound by two 1-s visual markers) and filled intervals (a continuous visual signal). The intervals were signaled by different visual stimuli, and they required responses to different sets of comparison stimuli. Empty intervals were judged as being longer than filled intervals. The difference between the point of subjective equality (PSE) for the empty intervals and the PSE for the filled intervals increased proportionally as the magnitudes of the anchor duration pairs were increased from 2 and 8 s to 4 and 16 s. In Experiment 2, the pigeons were trained to discriminate intervals signaled by the absence of houselight illumination (Group Empty) or the presence of houselight illumination (Group Filled). The psychophysical timing functions for these intervals were identical to each other. The results of Experiment 1 indicate that memory mixing is not necessary for detecting a timing difference between empty and filled intervals in pigeons. The results of Experiment 2 suggest that the nature of the stimuli that signal the empty and filled intervals impacts how pigeons judge the durations of empty and filled intervals.     

Testing the scalar expectancy theory (SET) and the learning-to-time model (LeT) in a double bisection task

Two theories of timing, scalar expectancy theory (SET) and learning-to-time (LeT), make substantially different assumptions about what animals learn in temporal tasks. In a test of these assumptions, pigeons learned two temporal discriminations. On Type 1 trials, they learned to choose a red key after a 1-sec signal and a green key after a 4-sec signal; on Type 2 trials, they learned to choose a blue key after a 4-sec signal and a yellow key after either an 8-sec signal (Group 8) or a 16-sec signal (Group 16). Then, the birds were exposed to signals 1 sec, 4 sec, and 16 sec in length and given a choice between novel key combinations (red or green vs. blue or yellow). The choice between the green key and the blue key was of particular significance because both keys were associated with the same 4-sec signal. Whereas SET predicted no effect of the test signal duration on choice, LeT predicted that preference for green would increase monotonically with the length of the signal but would do so faster for Group 8 than for Group 16. The results were consistent with LeT, but not with SET.     

Alterations in the memory code for temporal events induced by differential outcome expectancies in pigeons

The effect of differential outcome expectancies on memory for temporal and nontemporal information was examined. Pigeons were trained to match short (2-sec) and long (8-sec) sample durations to red and green comparison stimuli, and vertical and horizontal lines to vertical and horizontal comparison stimuli. In Experiment 1, one differential outcome (DO) group received food for correct choices on short-sample trials, whereas another received food for correct choices on long-sample trials. On line-orientation trials, half of each DO group received food for correct responses following vertical samples, whereas the other half received food for correct responses following horizontal samples. Overall retention was greater in the DO groups than in a nondifferential (NDO) group that received either food or no food for correct responses on a random half of all trials. Furthermore, although the NDO group displayed a choose-short bias for temporal samples, both DO groups displayed equivalent biases to select the comparison stimulus associated with food. In Experiment 2, differential outcome expectancies were extinguished off-baseline. Subsequently, in the first nondifferential outcome test session, the. DO groups performed less, accurately than the NDO group. These findings indicate that temporal samples are not retrospectively and analogically coded when they are differentially associated with food and no food. Instead, they are remembered in terms of the corresponding outcome expectancies.     

Stimulus duration and conditioned reinforcing value measured by a learning-tests procedure

The relationship between the duration of stimuli and their conditioned reinforcing effect was investigated using a learning-tests procedure. In Experiment 1, stimuli were the same duration on training (stimulus → reward) and test (choice response → stimulus). Ten- and 30-sec stimuli provided effective differential conditioned reinforcement but 3-sec stimuli did not. In Experiment 2, different pigeons had each combination of the 3- and 30-sec stimuli on training and test trials. Evidence of conditioned reinforcement was obtained only for the birds with 30-sec stimuli on both training and test. The results were interpreted as indicating that stimuli become effective conditioned reinforcers on test trials only when their duration exceeds the duration of differential short-term memory cues resulting from a difference in the events that precede them on training and test trials.     

Errors in pigeons’ memory for number of events

In Experiment 1, pigeons were trained in a within-subjects design to discriminate sequences of light flashes (illumination of the feeder) that varied in number, but not in time (2f/4sec and 8f/4sec), and in time, but not in number (4f/2sec and 4f/8sec). Number samples required a response to one of two comparison dimensions (either color or line), whereas time samples required a response to the remaining comparison dimension. Delay testing revealed a significant choose-small bias following number samples and a significant choose-long bias following time samples. In Experiment 2, testing confirmed that in the absence of a sample, there was a bias to respond small to the number comparisons and long to the time comparisons. Additional tests indicated that the birds were discriminating time samples on the basis of the number of light flashes occurring during the last few seconds of the time samples, rather than on the basis of the total duration of the flash sequence. Consequently, the choose-long bias observed for time samples during delay testing was really a choose-small bias. In Experiment 3, the birds received baseline training with a 5-sec delay and were subsequently tested at shorter and longer delays. A choose-large bias occurred at delays shorter than the baseline training delay, whereas a choose-small bias was again observed at delays longer than the baseline delay. These findings provide additional empirical support for the conceptualizing of memory for number and time in terms of a common mechanism.     

Differential trace conditioning to temporal compounds

Rats were reinforced for responding following the presentation of a light-tone temporal compound, each component having a .5-sec duration separated by a .5-sec empty interval. For different groups, nonreinforced presentations of the compound in the reverse order and either or both of the components separately were included in each session. Results indicated a progressive increase in difficulty as the number of nonreinforced events increased, with behavior being strongly affected by the first component of the compound. Nevertheless, differential performance under the most demanding conditions demonstrate the rat’s ability to acquire a temporal compound discrimination, which suggests an interpretation based on interacting stimulus traces.     

Temporal cuing of runs in series of reward events reduces interevent anticipation

In Experiment 1, a group of rats were runway trained on each of two reward series for 32 days. The two series consisted of three runs, the first two of which were, respectively, rewarded and nonrewarded; the third run was rewarded in one series but nonrewarded in the other. A 40-min interval separated the two series; the first and second runs within the series were separated by a 10-min interval, whereas the second and third runs were separated by a 30-sec interval. The reward (and nonreward) events and temporal cues of the two series are designated R-NR/R-NN. A second group was similarly trained, with the exception that the 10-min interval separated the second and third runs (RN-R/RN-N). Both groups developed appropriate differential running on the third run of the two series, and the RN-R/RN-N animals ran appropriately (slowly) on the second run of both series. Appropriate Run 2 performance appeared in one half of the R-NR/R-NN animals (depending upon order of series presentation); the remaining half ran faster on Run 2 of the R-NR series than on the same run of the R-NN series, an effect currently termed interevent anticipation. A cue shift phase in which all within-series intervals were 30 sec showed that the temporal intervals were controlling performance before the shift. Experiment 2 showed that interevent anticipation appears when all within-series intervals are either 10 min or 30 sec from the beginning of training, suggesting that the elimination of interevent anticipation in Experiment 1 was due to the differential cuing of runs by the temporal intervals rather than the particular interval duration. The overall findings suggest that the similarity of Run 2 and Run 3 performance termed interevent anticipation may be due to a failure to discriminate the ordinal position of runs within a series.     

Flexible coding of temporal information by pigeons: Event durations as remember and forget cues for temporal samples

The ability of pigeons to use event durations as remember (R) and forget (F) cues for temporal samples was examined. Pigeons were required to indicate whether a houselight sample stimulus was short (2 sec) or long (6 sec) by pecking a red or a green comparison stimulus. After training with a constant 10-sec delay interval, temporal cues (illumination of the center key) were presented 2 sec after the offset of the temporal samples. For one group, a short (2-sec) temporal cue served as the R cue and a long (6-3ec) temporal cue served as the F cue. This was reversed for a second group of birds. During training, comparison stimuli were always presented following the temporal R cue, but never following the temporal F cue. Tests for the effectiveness of the temporal R and F cues showed that F cues were equally effective in reducing matching accuracy in both groups of birds. It was concluded that pigeons used the duration of the cue to determine whether or not to rehearse the memory code for the temporal sample.     

Pigeons’ memory for sequences of light flashes when gap duration is an unreliable discriminative cue

In Experiment 1, pigeons were trained with a 1-sec dark and a 1-sec houselight-illuminated delay interval to discriminate between sequences of two and four flashes of light (feeder illumination). The sequences could be discriminated on the basis of the number of flashes, the number of gaps, or the duration of the gap between flashes. A choose-few bias was obtained at extended dark delays, but not at extended illuminated delays. Pigeons appeared to confuse long dark delays with the longer gap between flashes on few-sample trials. In Experiment 2, additional sample sequences were included that made gap duration an unreliable cue for discriminating between the few and many samples. A significant choose-many bias was obtained at extended dark delay intervals, but no biased forgetting was found at extended illuminated delays. The pigeons appeared to discriminate light flash sequences by relying on multiple temporal features of a sequence rather than using an event switch to count flashes. The biased-forgetting effects observed appear to be due to instructional ambiguity that results from the similarity of the delay interval to features of the flash sequences. nt]mis|This research was supported by Grant OGPOOD6378 from the Natural Sciences and Engineering Research Council of Canada to A.S.     

Interference and auditory short-term memory in the bottlenosed dolphin

Interference in auditory short-term memory in the bottlenosed dolphin,Tursiops truncatus (Montagu), was studied using a delayed matching-to-sample task. At each trial, one of two sample sounds, chosen randomly, was projected underwater for 4 sec and then, after a variable delay interval, both sounds were presented. A response to the sound matching the initial sample was reinforced. Correct matching was significantly reduced following short intervals between trials in combination with long delays after the sample (proactive interference), or when a near continuous irrelevant sound was inserted into the delay interval (retroactive interference). There was rapid habituation to interference if the irrelevant sound was short in duration relative to the delay interval. For both proactive and retroactive interference, the errors were predominantly responses to the sample sound appropriate to the prior trial rather than to the current trial, indicating that memory for the relative recency of events (temporal memory) was degraded by interference. When interference was deleted or minimized, temporal memory remained nearly perfect over 30-sec delay intervals, the longest tested. The importance of distinguishing between temporal memory and nontemporal, or event, memory in different forms of the delayed matching task was emphasized.     

Proactive effects in pigeons’ timing behavior: Implications for an internal-clock model

We have found proactive effects in pigeons’ timing behavior, a finding inconsistent with internal-clock models of timing that assume a resetable working-memory component. Six pigeons were trained to discriminate between 2- and 10-sec illuminations of a white light; choice of a red pecking key was correct and rewarded after presentation of the short stimulus whereas choice of a green key was correct and rewarded after presentation of the long stimulus. During training sessions, there were 60 trials separated by a 20-sec intertriai interval; short and long light occurred in a randomized order and correct choices were reinforced with 5-sec access to grain on a partial (75%) schedule. During test sessions, there were 120 trials separated by a 2-sec intertrial inter val. Light presentations occurred in a fixed order throughout these sessions: 2, 6, 10, 10, 6, 2 2, 6, 10 sec, and so forth. Choice of either red or green after 6 sec was not reinforced. However, red continued to be correct after 2 sec and green continued to be correct after 10 sec. Of central interest was how the subjects classified 6 sec of light in ascending (2, 6, 10) and descending (10. 6, 2) sequences of durations: Subjects chose the short alternative on 42% of the 6-sec trials in ascending series but only 29% in descending series, a result most plausibly interpreted as show ing that duration information from a preceding trial affects duration classifications on the cur rent trial. Such proactive effects should not occur according to working-memory models that as sume that stored information is cleared at the end of a trial.     

Absence of immediate transfer of training of duration symbolic-matching-to-sample in pigeons

Two experiments examined the performance of pigeons on symbolic-matching-to sample in which the relevant sample dimension consisted of duration. Each pigeon was trained on two problems that had the same two sample durations, 2 and 10 sec, but were different with respect to other physical properties of the samples. Durations of light and tone were used in Experiment 1; durations of two different color-location compounds were used in Experiment 2. In each experiment, a unique choice stimulus was associated with each of the four possible combinations of duration and signal type. Test sessions contained probe trials in which the choice stimuli were these appropriate for a long and a short duration of the signal type opposite to that actually presented. Pigeons in both experiments displayed asymmetrical performance deficits. Accuracy on long durations dropped to chance or below, whereas accuracy on short durations remained high. This pattern is similar to the choose-short effect that is obtained when animals are tested with long retention intervals. The implications of these results for duration memory, coding, and transfer of training are discussed.