Non-threatening Stimuli Research Articles

The effect of spatial frequency information and visual similarity in threat detection

ABSTRACTIn the current research, we sought to examine the role of spatial frequency on the detection of threat using a speeded visual search paradigm. Participants searched for threat-relevant (snakes or spiders) or non-threat-relevant (frogs or cockroaches) targets in an array of neutral (flowers or mushrooms) distracters, and we measured search performance with images filtered to contain different levels (high and low) of spatial frequency information. The results replicate previous work demonstrating more rapid detection of threatening versus non-threatening stimuli [e.g. LoBue, V. & DeLoache, J. S. (2008). Detecting the snake in the grass: Attention to fear-relevant stimuli by adults and young children. Psychological Science, 19, 284–289. doi:10.1111/j.1467-9280.2008.02081.x]. Most importantly, the results suggest that low spatial frequency or relatively coarse levels of visual information is sufficient for the rapid and accurate detection of threatening stimuli. Furthermore, the results also suggest that visual similarity between the stimuli used in the search tasks plays a significant role in speeded detection. The results are discussed in terms of the theoretical implications for the rapid detection of threat and methodological implications for properly accounting for similarity between the stimuli in visual search studies.

Does Seeing Faces of Young Black Boys Facilitate the Identification of Threatening Stimuli?

Pervasive stereotypes linking Black men with violence and criminality can lead to implicit cognitive biases, including the misidentification of harmless objects as weapons. In four experiments, we investigated whether these biases extend even to young Black boys (5-year-olds). White participants completed sequential priming tasks in which they categorized threatening and nonthreatening objects and words after brief presentations of faces of various races (Black and White) and ages (children and adults). Results consistently revealed that participants had less difficulty (i.e., faster response times, fewer errors) identifying threatening stimuli and more difficulty identifying nonthreatening stimuli after seeing Black faces than after seeing White faces, and this racial bias was equally strong following adult and child faces. Process-dissociation-procedure analyses further revealed that these effects were driven entirely by automatic (i.e., unintentional) racial biases. The collective findings suggest that the perceived threat commonly associated with Black men may generalize even to young Black boys.

Uncertainty is associated with increased selective attention and sustained stimulus processing.

Uncertainty about future threat has been found to be associated with an overestimation of threat probability and is hypothesized to elicit additional allocation of attention. We used event-related potentials to examine uncertainty-related dynamics in attentional allocation, exploiting brain potentials' high temporal resolution and sensitivity to attention. Thirty participants performed a picture-viewing task in which cues indicated the subsequent picture valence. A certain-neutral and a certain-aversive cue accurately predicted subsequent picture valence, whereas an uncertain cue did not. Participants overestimated the effective frequency of aversive pictures following the uncertain cue, both during and after the task, signifying expectancy and covariation biases, and they tended to express lower subjective valences for aversive pictures presented after the uncertain cue. Pictures elicited increased P2 and LPP amplitudes when their valence could not be predicted from the cue. For the LPP, this effect was more pronounced in response to neutral pictures. Uncertainty appears to enhance the engagement of early phasic and sustained attention for uncertainly cued targets. Thus, defensive motivation related to uncertainty about future threat elicits specific attentional dynamics implicating prioritization at various processing stages, especially for nonthreatening stimuli that tend to violate expectations.

A preliminary investigation into psychophysiological effects of threatening a perceptually embodied rubber hand in healthy human participants

Background and aimsThreatening a perceptually embodied rubber hand with noxious stimuli has been shown to generate levels of anxiety similar to that experienced when a real hand is threatened. The aim of this study was to investigate skin conductance response, self-reported anxiety and the incidence, type and location of sensations when a perceptually embodied rubber hand was exposed to threatening and non-threatening stimuli. MethodsA repeated measures cross-over design was used whereby 20 participants (≥18 years, 14 females) received a threatening (syringe needle) and non-threatening (soft brush) stimulus to a perceptually embodied rubber hand. Perceptual embodiment was achieved using a soft brush to synchronously stroke the participant's real hand (out of view) and a rubber hand (in view). Then the investigator approached the rubber hand with a syringe needle (threat) or soft brush (non-threat). ResultsRepeated measures ANOVA found that approaching the perceptually embodied rubber hand with either stimulus produced statistically significant reductions in the rated intensity of response to the following questions (p<0.01): ‘How strongly does it feel like the rubber hand is yours?’; ‘How strongly does it feel like the rubber hand is part of your body?’; and ‘How strongly does it feel you can move the rubber hand?’. However, there were no statistically significant differences in scores between needle and brush stimuli. Repeated measures ANOVA on skin conductance response found statistically significant effects for experimental Events (baseline; stroking; perceptual embodiment; stimuli approaching rubber hand; stimuli touching rubber hand; p<0.001) but not for Condition (needle versus brush p=0.964) or experimental Event×Condition interaction (p=0.160). Ten of the 20 participants (50%) reported that they experienced a sensation arising from the rubber hand when the rubber hand was approached and touched by either the needle and/or brush but these sensations lacked precision in location, timing, and nature. Conclusion and implicationsOur preliminary findings suggest that the increase in arousal in response to stimuli entering the peripersonal space may not be selective for threat. There was tentative evidence that more intense sensations were experienced when a perceptually embodied rubber hand was approached by a threatening stimulus. Our findings provide initial insights and should serve as a catalyst for further research.

Distracted by danger: Temporal and spatial dynamics of visual selection in the presence of threat.

Threatening stimuli are known to influence attentional and visual processes in order to prioritize selection. For example, previous research showed faster detection of threatening relative to nonthreatening stimuli. This has led to the proposal that threatening stimuli are prioritized automatically via a rapid subcortical route. However, in most studies, the threatening stimulus is always to some extent task relevant. Therefore, it is still unclear if threatening stimuli are automatically prioritized by the visual system. We used the additional singleton paradigm with task-irrelevant fear-conditioned distractors (CS+ and CS-) and indexed the time course of eye movement behavior. The results demonstrate automatic prioritization of threat. First, mean latency of saccades directed to the neutral target was increased in the presence of a threatening (CS+) relative to a nonthreatening distractor (CS-), indicating exogenous attentional capture and delayed disengagement of covert attention. Second, more error saccades were directed to the threatening than to the nonthreatening distractor, indicating a modulation of automatically driven saccades. Nevertheless, cumulative distributions of the saccade latencies showed no modulation of threat for the fastest goal-driven saccades, and threat did not affect the latency of the error saccades to the distractors. Together these results suggest that threatening stimuli are automatically prioritized in attentional and visual selection but not via faster processing. Rather, we suggest that prioritization results from an enhanced representation of the threatening stimulus in the oculomotor system, which drives attentional and visual selection. The current findings are interpreted in terms of a neurobiological model of saccade programming.

Threat modulates neural responses to looming visual stimuli.

Objects on a collision course with an observer produce a specific pattern of optical expansion on the retina known as looming, which in theory exactly specifies the time-to-collision (TTC) of approaching objects. It was recently demonstrated that the affective content of looming stimuli influences perceived TTC, with threatening objects judged as approaching sooner than non-threatening objects. Here, the neural mechanisms by which perceived threat modulates spatiotemporal perception were investigated. Participants judged the TTC of threatening (snakes, spiders) or non-threatening (butterflies, rabbits) stimuli, which expanded in size at a rate indicating one of five TTCs. Visual-evoked potentials (VEPs) and oscillatory neural responses measured with electroencephalography were analysed. The arrival time of threatening stimuli was underestimated compared with non-threatening stimuli, though an interaction suggested that this underestimation was not constant across TTCs. Further, both speed of approach and threat modulated both VEPs and oscillatory responses. Speed of approach modulated the N1 parietal and oscillations in the beta band. Threat modulated several VEP components (P1, N1 frontal, N1 occipital, early posterior negativity and late positive potential) and oscillations in the alpha and high gamma band. The results for the high gamma band suggest an interaction between these two factors. Previous evidence suggests that looming stimuli activate sensorimotor areas, even in the absence of an intended action. The current results show that threat disrupts the synchronization over the sensorimotor areas that are likely activated by the presentation of a looming stimulus.

Autonomic arousal and attentional orienting to visual threat are predicted by awareness.

The rapid detection and evaluation of threat is of fundamental importance for survival. Theories suggest that this evolutionary pressure has driven functional adaptations in a specialized visual pathway that evaluates threat independently of conscious awareness. This is supported by evidence that threat-relevant stimuli rendered invisible by backward masking can induce physiological fear responses and modulate spatial attention. The validity of these findings has since been questioned by research using stringent, objective measures of awareness. Here, we use a modified continuous flash suppression paradigm to ask whether threatening images induce adaptive changes in autonomic arousal, attention, or perception when presented outside of awareness. In trials where stimuli broke suppression to become visible, threatening stimuli induced a significantly larger skin conductance response than nonthreatening stimuli and attracted spatial attention over scrambled images. However, these effects were eliminated in trials where observers were unaware of the stimuli. In addition, concurrent behavioral data provided no evidence that threatening images gained prioritized access to awareness. Taken together, our data suggest that the evaluation and spatial detection of visual threat are predicted by awareness.

Multimodal processing of emotional information in 9-month-old infants I: Emotional faces and voices

Making sense of emotions manifesting in human voice is an important social skill which is influenced by emotions in other modalities, such as that of the corresponding face. Although processing emotional information from voices and faces simultaneously has been studied in adults, little is known about the neural mechanisms underlying the development of this ability in infancy. Here we investigated multimodal processing of fearful and happy face/voice pairs using event-related potential (ERP) measures in a group of 84 9-month-olds. Infants were presented with emotional vocalisations (fearful/happy) preceded by the same or a different facial expression (fearful/happy). The ERP data revealed that the processing of emotional information appearing in human voice was modulated by the emotional expression appearing on the corresponding face: Infants responded with larger auditory ERPs after fearful compared to happy facial primes. This finding suggests that infants dedicate more processing capacities to potentially threatening than to non-threatening stimuli.

Does attention bias modification improve attentional control? A double-blind randomized experiment with individuals with social anxiety disorder

People with anxiety disorders often exhibit an attentional bias for threat. Attention bias modification (ABM) procedure may reduce this bias, thereby diminishing anxiety symptoms. In ABM, participants respond to probes that reliably follow non-threatening stimuli (e.g., neutral faces) such that their attention is directed away from concurrently presented threatening stimuli (e.g., disgust faces). Early studies showed that ABM reduced anxiety more than control procedures lacking any contingency between valenced stimuli and probes. However, recent work suggests that no-contingency training and training toward threat cues can be as effective as ABM in reducing anxiety, implying that any training may increase executive control over attention, thereby helping people inhibit their anxious thoughts. Extending this work, we randomly assigned participants with DSM-IV diagnosed social anxiety disorder to either training toward non-threat (ABM), training toward threat, or no-contingency condition, and we used the attention network task (ANT) to assess all three components of attention. After two training sessions, subjects in all three conditions exhibited indistinguishably significant declines from baseline to post-training in self-report and behavioral measures of anxiety on an impromptu speech task. Moreover, all groups exhibited similarly significant improvements on the alerting and executive (but not orienting) components of attention. Implications for ABM research are discussed.

Comparison of inbred mouse substrains reveals segregation of maladaptive fear phenotypes.

Maladaptive fear, such as fear that is persistent or easily generalized to a nonthreatening stimuli, is associated with anxiety-related disorders in humans. In the laboratory, maladaptive fear can be modeled in rodents using Pavlovian fear conditioning. Recently, an inbred mouse strain known as 129S1/SvImJ, or 129S1 has been reported as exhibiting impairments in fear extinction and enhanced fear generalization. With a long-term goal of identifying segregating genetic markers of maladaptive fear, we used Pavlovian fear conditioning to characterize a closely related substrain designated as 129S6/SvEvTac, or 129S6. Here we report that, like 129S1 animals, 129S6 mice exhibit appropriate levels of fear upon conditioning, but are unable to extinguish fear memories once they are consolidated. Importantly, the maladaptive fear phenotype in this inbred stain can be segregated by sub-strain when probed using conditioning protocols designed to assess generalized fear. We find that unlike the 129S1 substrain, mice from the 129S6 sub-strain do not generalize conditioned fear to previously novel contexts and can learn to discriminate between two similar contexts when trained using a discrimination protocol. These results suggest that at least two forms of maladaptive fear (deficits in fear extinction and fear generalization) can be can be functionally segregated, further suggesting that the underlying neurobiology is heritable. Given the observation that two closely related sub-strains can exhibit different constellations of maladaptive fear suggests that these findings could be exploited to facilitate the identification of candidate genes for anxiety-related disorders.

Evolutionary preferences for physical formidability in leaders.

This research uses evolutionary theory to evaluate followers' preferences for physically formidable leaders and to identify conditions that stimulate those preferences. It employs a population-based survey experiment (N ≥ 760), which offers the advantages to internal validity of experiments and external validity of a highly heterogeneous sample drawn from a nationally representative subject pool. The theoretical argument proffered here is followers tend to prefer leaders with greater physical formidability because of evolutionary adaptations derived from humans' violent ancestral environment. In this environment, individuals who allied with and ultimately followed physically powerful partners were more likely to acquire and retain important resources necessary for survival and reproduction because the presence of the physically powerful partner cued opponents to avoid a challenge for the resources or risk a costly confrontation. This argument suggests and the results indicate that threatening (war) and nonthreatening (peace, cooperation, and control) stimuli differentially motivate preferences for physically formidable leaders. In particular, the findings suggest threatening conditions lead to preferences for leaders with more powerful physical attributes, both anthropometric (i.e., weight, height, and body mass index) and perceptual (i.e., attributes of being "physically imposing or intimidating" and "physically strong"). Overall, this research offers a theoretical framework from which to understand this otherwise seemingly irrational phenomenon. Further, it advances the emerging but long-neglected investigation of biological effects on political behavior and has implications for a fundamental process in democratic society, leader selection.

Bullets Versus Burgers: Is It Threat or Relevance That Captures Attention?

Previous studies have found that potentially dangerous stimuli are better at capturing attention than neutral stimuli, a finding sometimes called the threat superiority effect. However, non-threatening stimuli also capture attention in many studies of visual attention. In Experiment 1, the relevance superiority effect was tested with a visual search task comparing detection times for threatening stimuli (guns), pleasant but motivationally relevant stimuli (food), and neutral stimuli (flowers and chairs). Gun targets were detected more rapidly than both types of neutral targets, whereas food targets were detected more quickly than the neutral chair targets only. Guns were detected more rapidly than food. In Experiment 2, threatening targets (guns and snakes), pleasant but motivationally relevant targets (money and food), and neutral targets (trees and couches) were all presented with the same neutral distractors (cactus and pots) in order to control for the valence of the distractor stimulus across the three categories of target stimuli. Threatening and pleasant target categories facilitated attention relative to neutral targets. The results support the view that both threatening and pleasant pictures can be detected more rapidly than neutral targets.

Sexually dimorphic patterns of neural activity in response to juvenile social subjugation

After experiencing juvenile social subjugation (JSS), adult female rats display more severe depression- and anxiety-like behaviors than adult males, suggesting that JSS is encoded in a sex-specific manner. To test this hypothesis, prepubertal rats (P28–33) were subjected to 10 aggressive acts in ≤10min from an aggressive adult male, a 10min encounter with a non-aggressive adult male, or to 10min in an empty, clean cage (handled control) and were sacrificed one hour later. We then used unbiased stereology to estimate the total number and proportion of neurons immunoreactive for the immediate early gene product Fos bilaterally in the basolateral amygdala (BLA), the anterior and posterior subdivisions of the bed nucleus of the stria terminalis, and the paraventricular nucleus of the hypothalamus (PVN). Overall, females’ Fos responses were less selective than males’. The BLA in males displayed a selective Fos response to the non-aggressive male, whereas no such selectivity occurred in the BLA of females. Additionally, there were more neurons overall in the left BLA than the right and this lateralization was specific to males. The principal subdivision of the BST (BSTpr) in males responded selectively to JSS, whereas the BSTpr in females responded to both the non-aggressive and aggressive males. We also found that the regional volume and neuron number of the BSTpr is greater in males than in females. Finally, the PVN in males was, like the BLA, selective for the non-aggressive male, whereas none of the experiences elicited a selective response in females. The greater selectivity for non-threatening stimuli in males in three stress-responsive brain regions may be a clue as to why males are less susceptible to the anxiogenic effects of JSS.

ENDURING INFLUENCE OF EARLY TEMPERAMENT ON NEURAL MECHANISMS MEDIATING ATTENTION-EMOTION CONFLICT IN ADULTS

Behavioral inhibition, a temperament identified in early childhood, is often associated with dysregulated attention and affective processing, particularly in response to threat. Longitudinal studies find that the manifestation of perturbed attention and affective processing often dissipates with age. Yet, childhood behavioral inhibition continues to predict perturbed brain function into adulthood. This suggests that adults with childhood behavioral inhibition may engage compensatory processes to effectively regulate emotion-related attention. However, it is unknown whether perturbations in brain function reflect compensation for attention bias to emotional stimuli generally, or to threatening contexts more specifically. The present study tests these possibilities. Adults with and without a history of stable childhood behavioral inhibition completed an attention-control task in the context of threatening and nonthreatening stimuli while undergoing functional magnetic resonance imaging. Participants were asked to identify the gender of fearful (threatening) and happy (nonthreatening) faces, while ignoring both the face emotion and overlaid congruent (low attention control, LAC) or incongruent (high attention control, HAC) gender words. When fearful faces were present, adults with stable childhood behavioral inhibition exhibited more activity in striatum, cingulate, and dorsolateral prefrontal cortex for HAC trials compared with LAC trials, relative to those without behavioral inhibition. When happy faces were present, the opposite activation pattern emerged. No group differences in behavior were observed. Among adults, stable childhood behavioral inhibition predicts neural, but not behavioral, responding when attention control is engaged in discrete emotional contexts. This suggests a mechanism by which adults may compensate for the behavioral manifestation of threat-based attention biases.

Varying expectancies and attention bias in phobic and non-phobic individuals.

Phobic individuals display an attention bias to phobia-related information and biased expectancies regarding the likelihood of being faced with such stimuli. Notably, although attention and expectancy biases are core features in phobia and anxiety disorders, these biases have mostly been investigated separately and their causal impact has not been examined. We hypothesized that these biases might be causally related. Spider phobic and low spider fearful control participants performed a visual search task in which they specified whether the deviant animal in a search array was a spider or a bird. Shorter reaction times (RTs) for spiders than for birds in this task reflect an attention bias toward spiders. Participants' expectancies regarding the likelihood of these animals being the deviant in the search array were manipulated by presenting verbal cues. Phobics were characterized by a pronounced and persistent attention bias toward spiders; controls displayed slower RTs for birds than for spiders only when spider cues had been presented. More important, we found RTs for spider detections to be virtually unaffected by the expectancy cues in both groups, whereas RTs for bird detections showed a clear influence of the cues. Our results speak to the possibility that evolution has formed attentional systems that are specific to the detection of phylogenetically salient stimuli such as threatening animals; these systems may not be as penetrable to variations in (experimentally induced) expectancies as those systems that are used for the detection of non-threatening stimuli. In sum, our findings highlight the relation between expectancies and attention engagement in general. However, expectancies may play a greater role in attention engagement in safe environments than in threatening environments.

Threat modulates perception of looming visual stimuli

Among the most critical of visual functions is the detection of potentially hazardous or threatening aspects of the environment. For example, objects on a collision course with an observer must be quickly identified to allow sufficient time to prepare appropriate defensive or avoidant responses. Directly approaching objects produce a specific accelerating pattern of optical expansion, known as 'looming, which in theory exactly specifies time-to-collision independent of object size or distance. Such looming stimuli have been shown to trigger stereotyped defensive responses in both monkeys [1] and human infants [2]. Psychophysical results in adult participants have similarly suggested sensitivity to looming at early stages of visual processing [3]. Such findings indicate specialization of the visual system to detect and react to such 'looming' stimuli, and have contributed to the traditional view of looming as a purely optical cue to imminent collision [1]. Here, we investigated whether the semantic content of a looming visual stimulus affects perceived time-to-collision by manipulating its threat value. We show that time-to-collision is underestimated for threatening (snakes and spiders) compared to non-threatening (butterflies and rabbits) stimuli. Further, the magnitude of this effect is correlated with self-reported fear. Our results demonstrate affective modulation of the perception of looming stimuli, and suggest that emotion shapes basic aspects of visual perception.

Preferential processing of threatening facial expressions using the repetition blindness paradigm

Background: Neuroanatomical evidence suggests that the human brain has dedicated pathways to rapidly process threatening stimuli. This processing bias for threat was examined using the repetition blindness (RB) paradigm. RB (i.e., failure to report the second instance of an identical stimulus rapidly following the first) has been established for words, objects and faces but not, to date, facial expressions. Methods: 78 (Study 1) and 62 (Study 2) participants identified repeated and different, threatening and non-threatening emotional facial expressions in rapid serial visual presentation (RSVP) streams. Results: In Study 1, repeated facial expressions produced more RB than different expressions. RB was attenuated for threatening expressions. In Study 2, attenuation of RB for threatening expressions was replicated. Additionally, semantically related but non-identical threatening expressions reduced RB relative to non-threatening stimuli. Conclusions: These findings suggest that the threat bias is apparent in the temporal processing of facial expressions, and expands the RB paradigm by demonstrating that identical facial expressions are susceptible to the effect.

Threat modulates perception of looming visual stimuli

Objects on a collision course with an observer produce a specific pattern of optical expansion on the retina known as looming, which in theory exactly specifies time-to-collision. Looming stimuli produce stereotyped defensive responses in monkeys and human infants, indicating that the primate visual system is intrinsically tuned to interpret this stimulus as threatening. We investigated how emotional reactions to the semantic content of a looming stimulus affects perceived time-to-collision. We presented either threatening (snakes, spiders) or non-threatening (butterflies and rabbits) stimuli which expanded in size at a rate indicating one of five different times to contact (3.0, 3.5, 4.0, 4.5 or 5.0 s). After 1 s, the image disappeared and participants judged when it would have made contact with them continuing on the same trajectory. Consistent with previous findings, time-to-collision judgments systematically underestimated true arrival time. More importantly, our results showed that the magnitude of this underestimating significantly increased for threatening, compared to non-threatening, stimuli. Further, the magnitude of this increase was correlated with participants’ self-reported fear of snakes and spiders. Traditionally, looming has been interpreted as a purely optical cue to collision. Against that view, our results demonstrate that the semantic content of visual stimuli modulates perceived time-to-contact. Our results suggest that perceived threat led participants to use a larger margin of safety: underestimating time-to-collision errs on the side of allowing more time to prepare defensive responses (or flee). More generally, these results demonstrate that emotion has widespread effects on even very basic aspects of perception.

The Causal Role of Attentional Bias for Threat Cues in Social Anxiety: A Test on a Cyber-Ostracism Task

Cognitive models of social phobia postulate that attentional biases for threat play an important role in the maintenance of this disorder (e.g., Clark 2001). Consistent with this idea, studies have demonstrated that training social phobics to attend to non-threatening stimuli results in clinical benefits (Amir et al. in J Abnorm Psychol 117:860–868, 2008). However, no study has directly examined the causal status of selective attentional bias in social phobia. The present study explicitly investigated this issue. We used an experimental design similar to MacLeod et al. (J Abnorm Psychol 111:107–123, 2002), which involved two consecutive experimental phases: an attentional bias induction phase and a stress phase. During the attentional bias induction, participants completed modified versions of a dot-probe task; for half of the participants the task was designed to induce a biased attentional response for faces expressing disgust, for the other half, the task induced no bias. Then, all participants were exposed to a task inducing social rejection. Results indicate that the induction of an attentional bias for threatening information resulted in increased anxiety during social rejection. Implications for cognitive models of social phobia are discussed.

Learning about non-predators and safe places: the forgotten elements of risk assessment

A fundamental prerequisite for prey to avoid being captured is the ability to distinguish dangerous stimuli such as predators and risky habitats from non-dangerous stimuli such as non-predators and safe locations. Most research to date has focused on mechanisms allowing prey to learn to recognize risky stimuli. The paradox of learned predator recognition is that its remarkable efficiency leaves room for potentially costly mistakes if prey inadvertently learn to recognize non-predatory species as dangerous. Here, we pre-exposed embryonic woodfrogs, Rana sylvatica, to the odour of a tiger salamander, Ambystoma tigrinum, without risk reinforcement, and later try to teach the tadpoles to recognize the salamander, a red-bellied newt Cynops pyrrhogaster—a closely related amphibian, or a goldfish, Carassius auratus, as a predator. Tadpoles were then tested for their responses to salamander, newt or fish odour. Pre-exposure to salamander did not affect the ability of tadpoles to learn to recognize goldfish as a predator. However, the embryonic pre-exposure to salamanders inhibited the subsequent learning of salamanders as a potential predator, through a mechanism known as latent inhibition. The embryonic pre-exposure also prevented the learned recognition of novel newts, indicating complete generalization of non-predator recognition. This pattern does not match that of generalization of predator recognition, whereby species learning to recognize a novel predator do respond, but not as strongly, to novel species closely related to the known predator. The current paper discusses the costs of making recognition mistakes within the context of generalization of predators and dangerous habitats versus generalization of non-predators and safe habitats and highlights the asymmetry in which amphibians incorporate information related to safe versus risky cues in their decision-making. Mechanisms such as latent inhibition allow a variety of prey species to collect information about non-threatening stimuli, as early as during their embryonic development, and to use this information later in life to infer the danger level associated with the stimuli.