Assessing Attention in Category Learning by Animals

Defining Attention

One of William James’s (1890/1950) important contributions to psychology was his emphasis on the phenomenology of selectively attending to some stimuli while simultaneously ignoring others. Attention, he said, “is the taking possession by the mind, in clear and vivid form, of one out of what seem several simultaneously possible objects or trains of thought” (pp. 403–404). James’s definition was later behaviorally grounded by Herbert Spencer Jennings (1906/1976), for whom attention was not a conscious mental state; rather, “at the basis of attention lies objectively the phenomenon that the organism may react to only one stimulus even though other stimuli are present which would, if acting alone, likewise produce a response” (p. 330).

Although many operational definitions of attention have since been offered (Sutherland & Mackintosh, 1971), the most general and established may be that of Reynolds (1961). He proposed that “an organism attends to an aspect of the environment if independent variation or independent elimination of that aspect brings about variation in the organism’s behavior” (p. 203). Reynolds closely adhered to B. F. Skinner’s (1953) proposition that attention represents a controlling relation between a discriminative stimulus and a response.

Assessing Attention

Reynolds (1961) empirically assessed attention in his well-known demonstration of attention in the pigeon. He first trained two pigeons on a go/no-go discrimination task: They were trained to peck a key when it displayed a white triangle on a red background (the positive discriminative stimulus, S+), but not to peck the key when it displayed a white circle on a green background (the negative discriminative stimulus, S–). Reynolds later gave the pigeons postacquisition tests with the triangle, circle, red background, and green background separately and randomly displayed. One pigeon responded only to the triangle, whereas the second pigeon responded only to the red background (cf. Wilkie & Masson, 1976). Reynolds’s pioneering project thus showed

that a pigeon may attend to only one of several aspects of a discriminative stimulus. Every part of the environment that is present when a reinforced response occurs may not subsequently be an occasion for the emission of that response. (p. 208)

Postacquisition tests such as Reynolds’s have remained the predominant means of assessing attention in animals. Indeed, as we later document, when enhanced with contemporary computational modeling tools, such tests can still yield interesting and important results. Yet postacquisition tests have a notable limitation: They measure the consequence but not the process of deploying attention.

One way to assess attention during rather than after learning is to train animals with a multiple-necessary-cues (MNC) task (Soto & Wasserman, 2010). For this task, a set of several compound stimuli is created from cues lying along two or more dimensions. In the simplest case of two dimensions (A and B), two values along each dimension (A1 and A2, B1 and B2) are factorially combined, and only one of the four resulting compound stimuli is paired with reinforcement: A1B1+, A1B2–, A2B1–, A2B2–. Control by dimensions A and B can be continuously monitored by separately calculating two discrimination ratios:

( A 1 B 1 + A 1 B 2 ) ÷ ( A 1 B 1 + A 1 B 2 + A 2 B 1 + A 2 B 2 )

(1)

( A 1 B 1 + A 2 B 1 ) ÷ ( A 1 B 1 + A 1 B 2 + A 2 B 1 + A 2 B 2 )

(2)

Ratio 1 assesses control by dimension A, and Ratio 2 assesses control by dimension B. If dimension A were eventually to exert sole control over responding, then Ratio 1 (contrasting compounds containing A1 vs. A2) would rise from .50 to 1.00, whereas Ratio 2 (contrasting compounds containing B1 vs. B2) would remain near .50.

All of the research in our laboratory (using touch-screen technology; Gibson et al., 2004) has considerably expanded the MNC task. For example, Vyazovska et al. (2014) gave pigeons a go/no-go discrimination task reinforcing their pecks to only one of 16 different compound stimuli created from all possible combinations of two values along four visual dimensions: shape (circle/square), size (large/small), line orientation (horizontal/vertical), and brightness (dark/light). Pigeons readily mastered the task; their rates of responding to each of the 15 S–s eventually fell to less than 15% of their rate of responding to the sole S+. Converting these 16 rates to four discrimination ratios, as described earlier, indicated that pigeons’ responding was controlled by all four stimulus dimensions. Across all pigeons, learning rates were similar for the four dimensions, although individual birds exhibited slightly different speeds of learning the different dimensions. Also, learning was faster the more dimensions along which the S–s differed from the S+. This latter finding documents the added benefit arising from increasing perceptual disparities between the S–s and the S+.

A most interesting finding—and one in agreement with the familiar notions of selective attention and limited capacity (Kruschke & Johansen, 1999; Pashler, 1998)—was that there were clear attentional trade-offs among the four dimensions during learning. These trade-offs were typically drops in discrimination accuracy for previously learned dimensions when later-learned dimensions were being acquired (Teng et al., 2015).

It is important to appreciate that, as was true in Reynolds’s work, the compound visual stimuli we used involved interpretive intricacies: Most notably, such integral stimuli (i.e., stimuli for which it is not possible to perceive or process one dimension without perceiving or processing the other) permitted both elemental and configural cues to control behavior and attention (Soto & Wasserman, 2010). Computer modeling has revealed that, in the case of the MNC task, many theoretical accounts must hypothesize a unique configural cue associated with each compound stimulus in order to explain the full details of performance (Vyazovska et al., 2014). Other tasks involving separable stimulus elements are less subject to this intricacy.

Attention to the separable elements of compound visual stimuli is often monitored in humans by gaze direction, although looking at a discrete stimulus may not be sufficient for behavioral control by the stimulus. That said, compelling evidence from research on attention and categorization attests to the intimate interrelation between gaze direction and attention to relevant visual features in humans (Blair et al., 2009; Rehder & Hoffman, 2005a, 2005b).

Such monitoring of visual gaze appears to be particularly promising for realizing Bushnell’s (1998) hope of developing an effective process-specific assessment tool on which animal models of attention might be based. Rehder and Hoffman (2005a) reported that human research participants (a) learned to optimally allocate their attention to stimulus dimensions in the process of discriminating categories, (b) initially fixated on all of the dimensions, and (c) selectively directed their eye fixations toward the relevant dimensions after they had greatly reduced their categorization errors. Encouraged by these results, we began a series of investigations exploring peck tracking and visual category learning in pigeons, also using tasks involving separable stimulus elements.

Peck Tracking and Visual Category Learning

Concern with attentional precursors to successful discrimination behavior is not new (Dinsmoor, 1985). Spence (1940) deemed it vital that, if an animal is to learn a visual discrimination, then it must also engage in suitable receptor-orienting acts: The “animal must learn to orient and fixate its head and eyes so as to receive the . . . relevant stimulus aspects” (p. 277). To Spence (1950), “such learning is itself an active trial-and-error process” (p. 169).

In our research, we both temporally and spatially separated the animal’s stimulus-reception response from its final categorization response; doing so allowed us to isolate two important aspects of performance that would otherwise be conflated if the animal were to respond directly to the stimulus to be categorized—the customary behavioral measure. We called this task and monitoring system peck tracking.

In our initial peck-tracking project (Castro & Wasserman, 2014), we trained pigeons to classify stimuli from two different artificial visual categories (Fig. 1, left). Each category exemplar contained two relevant features (perfect predictors of category membership) and two irrelevant features. There were two relevant features for Category A and two different relevant features for Category B. The irrelevant features were common to both Categories A and B, so they could not predict category membership. Each of the relevant and irrelevant features appeared equally often in each of the four corner locations: top left, top right, bottom left, and bottom right. Thus, spatial location could not signal where the relevant features would be presented.

OPEN IN VIEWER

Fig. 1.

Stimuli and results from the first peck-tracking project (Castro & Wasserman, 2014). On the left are examples of the Category A and Category B training exemplars. There were two relevant features for Category A (the radiating color bands and the red bubbles) and two different relevant features for Category B (the green spiral and the Mondrian-colored squares). The other stimuli were irrelevant features, common to Categories A and B. On the right, the top graph shows the mean percentages of categorization accuracy and relevant pecks to stimulus features across training blocks in Experiment 1, and the bottom graph shows categorization accuracy across training blocks, separately for trials on which a relevant feature had been pecked just before making the choice response and trials on which an irrelevant feature had been pecked. Error bars indicate ±1 SEM.

When each exemplar was presented on the computer screen, pigeons had to peck it multiple times in order to move forward with the task. Critically, only the areas occupied by the relevant and irrelevant features were active for monitoring pecks; pecks at the black background did not count. But the pigeons were free to peck any of the features, relevant or irrelevant. We recorded the location of the pigeons’ pecks in order to determine whether or not they selectively directed their pecks to the relevant features of the category exemplars.

After the pigeons completed the required number of pecks at the stimulus, two report buttons appeared: one to the left and one to the right of the category exemplar. The birds had to peck one of the report buttons in order to classify the exemplar as belonging to Category A or Category B. Correct choice responses were followed by food reinforcement; incorrect choice responses were not followed by food, and correction trials continued until pigeons made the correct response.

As training proceeded, categorization accuracy increased, as expected. Moreover, the pigeons also increasingly pecked the relevant stimulus features (Fig. 1, top right), which suggests that they were tracking the relevant information for solving the task.

This parallel in performance prompted another question: Did attention to the relevant features rise first and promote category learning, or did categorization accuracy rise before the pigeons selectively attended to the relevant features? Rehder and Hoffman (2005a) found that people’s eye movements toward the relevant stimulus elements followed rather than preceded improvements in categorization accuracy; correct categorization responses were already frequent before people fully deployed attention to the relevant category features. In our case, two findings similarly suggested that pigeons’ tracking of the relevant stimuli followed increases in categorization accuracy.

First, the pigeons’ categorization accuracy was considerably higher after they pecked the relevant category features than after they pecked the irrelevant features (Fig. 1, bottom right). Second, using a state-space model (A. C. Smith et al., 2004), we mathematically confirmed that increases in peck tracking (i.e., increases in the rate of pecking relevant features) significantly followed increases in categorization accuracy (Castro & Wasserman, 2016). So, if pigeons make a correct choice response and consequently receive reinforcement, then they will pay increased attention to the previously chosen feature on future trials; attention then shifts to those features that prove to be reliable predictors of reinforcement, in accord with several prominent models of attentional learning (e.g., George & Pearce, 2012; Kruschke, 2001; Mackintosh, 1975).

Pigeons’ attention to relevant and irrelevant attributes of training exemplars can therefore be effectively monitored during category learning. Thus, peck tracking can serve as a useful measure of attention in pigeons, much as eye tracking is a useful measure of attention in humans (Castro & Wasserman, 2017; Sheridan et al., 2019).

Computational Modeling of Postacquisition Performance

Skinner (1953) argued that “attention is more than looking at something. . . . An organism is attending to a detail of a stimulus whether or not its receptors are oriented to produce the most clear-cut reception, if its behavior is predominantly under control of that detail” (p. 124). Although we have found that peck tracking can serve as a useful proxy for the on-line assessment of visual attention, we are cognizant that establishing the controlling relation between the details of a stimulus and the organism’s response always remains primary.

We have recently turned to computational modeling, having seen that it can be useful for making fresh inroads into assessing that controlling relation. This approach is also amenable to on-line assessment, although we have thus far limited our work to postacquisition assessment in order to validate its analytic utility.

In our peck-tracking experiments, one or more visual features were always perfect predictors of category membership. Under these circumstances, pigeons discovered the perfectly predictive features, and their attention was directed to those features (see also Lea et al., 2009; Wills et al., 2009). But it is important to appreciate that a category discrimination can also be accomplished by perceiving the overall similarity, or family resemblance, of the exemplars in each category; in this case, attention may be widely distributed among multiple features.

For example, Deng and Sloutsky (2016; see also L. B. Smith & Kemler, 1977) reported that infants learn the statistical co-occurrence of several features within exemplars of different categories, which suggests that their attention is distributed rather than focused on specific diagnostic features. What about pigeons’ attentional selectivity? Would they predominantly attend to a perfect predictor, as prior studies suggested? Or, if other features were also to have some lesser predictive value, would pigeons attend more diffusely, as young children do?

To find out, we gave human adults and pigeons a categorization task in which there was a single rule-like deterministic feature that perfectly predicted category membership accompanied by six other features that only probabilistically predicted category membership (Castro et al., 2020). Thus, the task could be learned on the basis of either one deterministic feature (which would encourage selective, focused attention) or multiple probabilistic features (which would encourage distributed attention).

Overall accuracy scores suggested that both humans and pigeons relied on the deterministic feature to categorize the stimuli, although humans may have done so to a greater degree. To gain a clearer understanding of humans’ and pigeons’ attention and categorization behavior, we tested them with new exemplars in which the deterministic feature was absent but the probabilistic features were present, as well as with new exemplars in which the deterministic feature was present but one or more of the probabilistic features were absent. Then, we used a modeling approach to determine subjects’ attentional profiles during testing.

We used Nosofsky’s (1986) generalized context model, a computational model which assumes that organisms represent categories by storing individual training exemplars in memory. According to this model, later classification of novel testing exemplars is based on similarity comparisons between the novel exemplars and the stored exemplars. This estimation of similarity is context dependent and influenced by the features to which the organism attends. Attention allocated to particular features—attentional weights—changes during training. Thus, we used the model to estimate the attentional weights that best accounted for the responses of humans and pigeons on categorization test trials.

We computed each subject’s attentional weight for each feature and then summarized each subject’s attentional profile by calculating the entropy of those attentional weights. Because we were interested in determining whether our subjects focused on one feature or distributed their attention to some or all of them, we deemed that entropy, a measure of variety or diversity provided by information theory (Shannon & Weaver, 1949), was a good candidate for this purpose. In our case, entropy measured the amount of informational diversity as a weighted average of the amount of information that each of the features in an exemplar provided. If only one feature carried all of the information (w = {1, 0, 0, 0, 0, 0, 0}), there was no informational diversity, so entropy was 0. If some or all of the features carried some amount of information, entropy was larger. Entropy was maximal if all of the features carried equal amounts of information (w = {.14, .14, .14, .14, .14, .14, .14}).

Once the entropy score for each subject was obtained, it was normalized on the basis of the maximum possible entropy given the number of features. Thus, normalized entropy was bounded by .00 and 1.00; 0.00 represented maximal selectivity (when the attentional weight for one of the features equaled 1.00, whereas the remaining weights for all of the other features equaled .00), and 1.00 represented minimal selectivity, or maximal distribution of attention (when all of the attentional weights were equal).

Most of the adults exhibited maximal selectivity—fully focusing on the deterministic feature—whereas the pigeons distributed their attention among several features (Fig. 2). Therefore, under these training conditions, pigeons do not appear to strongly focus attention on deterministic information and filter out less predictive information, as do human adults. Pigeons, as do young children (Deng & Sloutsky, 2016; L. B. Smith & Kemler, 1977), tend to distribute their attention among various elements of a stimulus.

OPEN IN VIEWER

Fig. 2.

Normalized entropy of the generalized context model’s best-fitting attentional weights for each individual subject in Castro et al. (2020). A value of 0 represents maximal selectivity (when the attentional weight for one of the features equals 1 and the remaining weights for all other features are 0), whereas a value of 1 represents minimal selectivity, or maximal distribution of attention (when all weights are equal). The dark blue points indicate the mean of each of the distributions (the error bars represent ±1 SEM). The dashed line, at .50, represents the middle value and is included simply as a reference.

Our peck-tracking experiments (Castro & Wasserman, 2014, 2016, 2017; Sheridan et al., 2019) indicated that pigeons selectively attend to the relevant features of category exemplars. In apparent contrast, the more recent study (Castro et al., 2020) indicated that pigeons distribute their attention among multiple stimulus features. Reconciling these apparently contradictory conclusions requires considering the predictive values of the stimulus features in the different studies. In the peck-tracking experiments, the relevant features had perfect predictive value, whereas the irrelevant features had no predictive value. In the more recent work, the deterministic feature was a perfect category predictor, but each of the probabilistic features also predicted category membership, albeit imperfectly (each was associated with the correct category response on 66% of the trials). It is thus possible that pigeons can learn to ignore entirely irrelevant features, but have more difficulty disengaging from features that have some predictive value but less than that of other, perfectly predictive features.

Final Remarks

Pigeons’ pecking behavior can provide important information about the process and consequences of visual attention. Touch-screen technology has allowed us to see that during learning, pigeons track those features that allow them to successfully solve a challenging categorization task. In this way, researchers can measure the ongoing process of attention in visual category learning. Critically, strong empirical parallels between pigeons and humans reveal that peck tracking can be a worthwhile measure of pigeons’ visual attention, much as eye tracking is considered a worthwhile measure of visual attention in human category learning (Rehder & Hoffman, 2005b). In addition, computational modeling techniques facilitate a better understanding of the consequences of visual attention even after learning has taken place.

These experimental strategies and computational methods facilitate and encourage a truly comparative study of the interplay between attention and categorization in humans and animals. Comparing diverse species should offer substantial insights into the evolutionary roots of categorization, specifically, and the nature of cognition, generally.

Recommended Reading

Güntürkün, O., Koenen, C., Iovine, F., Garland, A., & Pusch, R. (2018). The neuroscience of perceptual categorization in pigeons: A mechanistic hypothesis. Learning & Behavior, 46(3), 229–241. https://doi.org/10.3758/s13420-018-0321-6. A recent review focused on the neurobiological foundations of visual categorization learning in pigeons and other avian species.

Rehder, B., & Hoffman, A. B. (2005b). (See References). A comprehensive evaluation of different results in attention and category learning in humans, using eye tracking and computational modeling techniques.

Washburn, D. A., & Taglialatela, L. A. (2012). The competition for attention in humans and other animals. In T. R. Zentall & E. A. Wasserman (Eds.), Oxford handbook of comparative cognition (pp. 100–116). Oxford University Press. https://doi.org/10.1093/oxfordhb/9780195392661.013.0007. A clear and wide-ranging review of different aspects of attention in human and nonhuman animals.