Artificial grammar learning (AGL) provides a useful tool for exploring rule learning strategies linked to general purpose pattern perception. To be able to directly compare performance of humans with other species with different memory capacities, we developed an AGL task in the visual domain. Presenting entire visual patterns simultaneously instead of sequentially minimizes the amount of required working memory. This approach allowed us to evaluate performance levels of two bird species, kea (Nestor notabilis) and pigeons (Columba livia), in direct comparison to human participants. After being trained to discriminate between two types of visual patterns generated by rules at different levels of computational complexity and presented on a computer screen, birds and humans received further training with a series of novel stimuli that followed the same rules, but differed in various visual features from the training stimuli. Most avian and all human subjects continued to perform well above chance during this initial generalization phase, suggesting that they were able to generalize learned rules to novel stimuli. However, detailed testing with stimuli that violated the intended rules regarding the exact number of stimulus elements indicates that neither bird species was able to successfully acquire the intended pattern rule. Our data suggest that, in contrast to humans, these birds were unable to master a simple rule above the finite-state level, even with simultaneous item presentation and despite intensive training. One contribution of 13 to a Theme Issue 'Pattern perception and computational complexity'.

Formal language theory has been extended to two-dimensional patterns, but little is known about two-dimensional pattern perception. We first examined spontaneous two-dimensional visual pattern production by humans, gathered using a novel touch screen approach. Both spontaneous creative production and subsequent aesthetic ratings show that humans prefer ordered, symmetrical patterns over random patterns. We then further explored pattern-parsing abilities in different human groups, and compared them with pigeons. We generated visual plane patterns based on rules varying in complexity. All human groups tested, including children and individuals diagnosed with autism spectrum disorder (ASD), were able to detect violations of all production rules tested. Our ASD participants detected pattern violations with the same speed and accuracy as matched controls. Children's ability to detect violations of a relatively complex rotational rule correlated with age, whereas their ability to detect violations of a simple translational rule did not. By contrast, even with extensive training, pigeons were unable to detect orientation-based structural violations, suggesting that, unlike humans, they did not learn the underlying structural rules. Visual two-dimensional patterns offer a promising new formally-grounded way to investigate pattern production and perception in general, widely applicable across species and age groups.

No abstract

Graphical abstract

Whether pattern-parsing mechanisms are specific to language or apply across multiple cognitive domains remains unresolved. Formal language theory provides a mathematical framework for classifying pattern-generating rule sets (or “grammars”) according to complexity. This framework applies to patterns at any level of complexity, stretching from simple sequences, to highly complex tree-like or net-like structures, to any Turing-computable set of strings. Here, we explored human pattern-processing capabilities in the visual domain by generating abstract visual sequences made up of abstract tiles differing in form and color. We constructed different sets of sequences, using artificial “grammars” (rule sets) at three key complexity levels. Because human linguistic syntax is classed as “mildly context-sensitive,” we specifically included a visual grammar at this complexity level. Acquisition of these three grammars was tested in an artificial grammar-learning paradigm: after exposure to a set of well-formed strings, participants were asked to discriminate novel grammatical patterns from non-grammatical patterns. Participants successfully acquired all three grammars after only minutes of exposure, correctly generalizing to novel stimuli and to novel stimulus lengths. A Bayesian analysis excluded multiple alternative hypotheses and shows that the success in rule acquisition applies both at the group level and for most participants analyzed individually. These experimental results demonstrate rapid pattern learning for abstract visual patterns, extending to the mildly context-sensitive level characterizing language. We suggest that a formal equivalence of processing at the mildly context sensitive level in the visual and linguistic domains implies that cognitive mechanisms with the computational power to process linguistic syntax are not specific to the domain of language, but extend to abstract visual patterns with no meaning.

We investigated the role of local and global context on visual patterns produced by normal participants, examining the effects of both top-down context (framing) and bottom-up content (element-internal symmetry) in a computer-based experimental framework. In the first study, we allowed participants to generate rectangles of arbitrary proportions and found an effect of framing on width-to-height ratios of rectangles produced, demonstrating the importance of taking visual framing into account when discussing hunman shape preferences. In a second study, using FlexTiles, an interactive pattern-generation framework, we showed that the patterns humans produce are influenced by local symmetrical properties of pattem elements. Participants also had to indicate preferences between pairs of pattem variants. We found that in some cases, pattem preferences and pattem production lead to different results. We conclude that visual context, either in the form of visual framing or local symmetries, changes aesthetic patterns that humans produce and prefer in predictable ways. These differences between the productive and perceptual preferences highlight the importance of using multiple methods when studying the human aesthetic sense.

It is often reported that people like objects that are in order, predictable, or can be processed fluently. We suggested that we do not enjoy simple order as much as the process of ordering and therefore we like images that allow for insight-even in perceptually challenging contexts. Furthermore, perceptual challenge together with a promise of perceptual ordering could trigger interest. We report two studies which utilized patterns produced with different intentions (to be liked or interesting, etc.) by rotating visual elements via the software Flextiles. Participants evaluated the patterns on various dimensions regarding the potential for order-detection and perceptual challenge. Liking was predictable by potential for order-detection but not by complexity. Meanwhile, interest was predictable by a moderate potential for order-detection together with high complexity. Furthermore, patterns intended to be interesting were associated with perceptual challenge: less obvious order, more flaws of order, and more time to decide whether the image contains an order. Study 2 additionally included patterns intended to be beautiful or ugly, as well as random patterns. Liking was again predictable by potential for order-detection. Interest, in contrast, was predictable by a combination of potential for order-detection and high complexity. Complexity alone was not a significant predictor of interest this time, whereas patterns intended to be interesting were more perceptually challenging than those intended to be liked or beautiful. Our findings indicate that liking might be related to the potential for ordering, but interest requires association with order but also perceptual challenge.

Visual patterns are a key phenomenon in human aesthetics, reflecting a human "sense of order" (Gombrich, 1984). Social effects on the producer of visual aesthetic output may shed light on intuitive aesthetic knowledge that laypeople can utilize without explicit instructions, with implications for the evolution of aesthetics in humans more generally. We apply all 3 methods suggested by Gustav Fechner (preference, production, and use; Fechner 1871, 1876) to visual geometrical patterns, showing that symmetrical patterns are not only used most frequently in real life but are also produced spontaneously in the lab and are rated as significantly more attractive than are random patterns. We demonstrate that an anticipated audience affects the pleasingness of geometrical patterns produced by participants in the laboratory. Visual patterns created by participants instructed to make patterns that others will like are indeed rated more highly than are those that were made under the instruction to make patterns that are mainly pleasing to the producer.