Search

Pigeons are susceptible to several size and length illusions, but in some cases the bias has been shown to be opposite to that seen in humans. To further investigate how their perceptual system works, we asked how pigeons perceive orientation illusions. We used the Zollner illusion, in which parallel lines look non-parallel due to series of short crosshatches superimposed on the lines. First, we trained six birds to peck at the narrower (or wider) of the two gaps at the end of a pair of non-parallel target lines. After adapting the subjects to target lines with randomly oriented crosshatches (which result in no illusion at least to humans), we tested the pigeons' responses on randomly inserted probe trials, in which crosshatches that should induce the standard Zollner-like illusion for humans replaced the random-oriented ones. The results suggested that pigeons do perceive an illusion from Zollner figures, but in the direction opposite to that of humans. We propose that pigeons, contrary to humans, may assimilate the two lines of different orientations (each main line and crosshatch), which results in underestimation of acute angles, and this in turn may lead to a reversed Zollner illusion. Such assimilation dominance appears consistent with previous reports obtained for line length and size illusions in this species. (Contains 4 figures.)

Whereas many mammals (some primates and mice) experience amodal completion, previous data split for avian species. However, experimental procedures have been quite different among the species, and thus a direct comparison of various avian species in the same experimental situation is needed. We tested whether bantams (Gallus gallus domesticus) would complete partly occluded figures using a visual search task on the touch monitor, which was successfully used in our previous study with pigeons. First, we trained 3 participants to search for a notched red diamond (a target) among complete diamonds (distracters). Next, white squares accompanied each figure with a small gap of a fixed size. In test, the location of the accompanying white squares sometimes changed. In some trials, the white squares exactly covered, or 'occluded,' the notched portion of the target. Humans are known to have great difficulty in finding such targets due to 'automatic' completion of the notched part. However, bantams met no such difficulty at all. This result and the demonstration by Forkman (1998) of hens' amodal completion of figures placed on a perspective background, suggest that the perspective cue may have an important role in amodal completion in this species. Keywords: amodal completion, perceptual grouping, visual search, bantams DOI: 10.1037/a0019459

Nakamura, Fujita, Ushitani, & Miyata (2006) have shown that pigeons perceive the standard Muller-Lyer illusion. In this report, the authors examined effects of bracket sizes on perception of this illusion in pigeons (Columba livia) and humans (Homo sapiens). In Experiment 1, three pigeons were retrained to classify six lengths of target lines into 'long' and 'short' by pecking two keys on the monitor, ignoring the two brackets oriented toward the same direction. In the tests that followed, the standard Muller-Lyer figures of different bracket sizes were presented. All birds chose 'long' more frequently for the figures having inward-pointing brackets (>), regardless of bracket sizes. The overestimation of the target lines of inward-pointing figures continued to increase in pigeons, whereas it decreased as the bracket size became longer in humans (Experiment 2). The results suggest that these two species perceive the standard Muller-Lyer illusion with long brackets in different ways. Perhaps pigeons might not perceive illusions induced by contrast with the surrounding stimuli. Keywords: standard Muller-Lyer illusion, bracket size, pigeons, humans

Nakamura, et al. recently showed that pigeons experience the standard Muller-Lyer illusion but not the reversed illusion induced by detaching the arrowheads from the target line. This study re-examined pigeons' perception of this reversed figure by using the stimuli known to induce the maximal contrast effect in humans (Fellows, 1967). Pigeons were retrained to classify six lengths of target lines into 'long' and 'short' categories by pecking two keys on the monitor, ignoring the two brackets so placed that these would not induce an illusion. In the test that followed, two birds' responses were not affected by directions of arrowheads, as shown in the previous study. The third pigeon significantly chose 'long' for inward-pointing brackets figures (> ), that is, the direction of illusion was reversed from what is expected in humans. These results suggest that pigeons may not experience illusions induced by contrast with the surrounding stimuli.

A target circle surrounded by larger 'inducer' circles looks smaller, and one surrounded by smaller circles looks larger than they really are. This is the Ebbinghaus--Titchener illusion, which remains one of the strongest and most robust of contrast illusions. Although there have been many studies on this illusion in humans, virtually none have addressed how nonhuman animals perceive the same figures. Here the authors show that the Ebbinghaus--Titchener figures also induce a strong illusion in pigeons but, surprisingly, in the other direction; that is, all five successfully trained pigeons judged the target circle surrounded by larger circles to be larger than it really is and vice versa. Further analyses proved that neither the gaps between target and inducer circles nor the cumulative weighted surface of these figural elements could account for the birds' responses. Pigeons are known to show similarities to humans on various cognitive and perceptual tasks including concept formation, short-term memory, and some visual illusions. Our results, taken together with pigeons' previously demonstrated failure at visual completion, provide strong evidence that pigeons may actually experience a visual world too different for us to imagine. Keywords: Ebbinghaus--Titchener circles, assimilation illusion, contrast illusion, Columba livia, Homo sapiens

Despite their impressive cognitive abilities, avian species have shown less evidence for metacognition than mammals. We suspect that commonly used tasks such as matching to sample might be too demanding to allow metacognitive processing within birds’ working memory. Here, we examined whether pigeons could control their behavior as a function of knowledge levels on a three-item sequence learning task, a reference memory task supposedly requiring fewer working memory resources. The experiment used two types of lists differing in familiarity. One was familiar to the pigeons through repeated exposure, whereas the other was novel in every new session. In test sessions, pigeons could choose between a trial with a hint specifying the next item to peck and one with no hint. However, successful responses in trials with a hint resulted in lowered rates of primary reinforcement:.60 in the first test and.75 in the second. Results showed that two of four pigeons chose the trial with a hint significantly more often before receiving a novel list than the familiar list in the four sessions of the first test, and three did so in the second test. Impressively, one bird showed robust evidence in the very first sessions in both tests. These results suggest that pigeons may monitor their long-term knowledge states and thereby control their environment before starting to solve a task. [ABSTRACT FROM AUTHOR]