To successfully interact in social contexts we need to react to communication signals, including emotional signals expressed by others. Observers can perceive emotions from others’ facial, bodily, vocal, and verbal expressions, and this perception is further informed by contextual information. Among the available emotional cues, non-verbal vocalizations represent powerful primitive signals (e.g., Cowen et al., 2019) that prompt reactions (e.g., Warren et al., 2006), but whose valence may not be so obvious when no other contextual information is available (Atias et al., 2019). Indeed, Atias et al. (2019) showed that participants could not distinguish the valence of human non-verbal vocalizations occurring in real intense positive (e.g., reacting to the reunion with a loved one) and negative (e.g., reacting to an attacker invading one’s home) situations when these vocalizations were presented alone. However, when the affective vocalizations were matched with the corresponding visual contexts, participants immediately understood their valence correctly. As another example, Lavan et al. (2015) demonstrated that facial expressions drive the emotional interpretation of intense laughter and crying, further supporting the importance of information drawn from visual contexts.

Although we are usually presented with “contextualized” vocalizations, this may not be the case of a blind person who cannot rely on visual cues such as facial expressions or gestures in interpreting vocal emotional signals. Consistent experimental evidence suggests that blind people compensate for the lack of visual input and experience by increased sensitivity of their tactile (e.g., Cattaneo et al., 2010a; Bauer et al., 2015; Gurtubay-Antolin and Rodríguez-Fornells, 2017) and auditory senses (for a recent review, Sabourin et al., 2022). For instance, in the auditory domain blind individuals outperform sighted controls in the processing of pitch changes (e.g., Arnaud et al., 2018), localization (e.g., Battal et al., 2020), voice recognition (Hölig et al., 2014, 2015), as well as in auditorily presented sentence comprehension (Loiotile et al., 2020). Available findings on the processing of vocal emotions in blind individuals are mixed, with some studies reporting enhanced discrimination of emotional vocalizations in blind participants compared to the sighted (Klinge et al., 2010), others reporting no differences (Gamond et al., 2017) or even worse performance in the blind (e.g., Martins et al., 2019; Chen and Whitney, 2022; Sarzedas et al., 2024). Interestingly, blindness has also been associated with physiological differences in the way emotional vocalizations are processed: Klinge et al. (2010) reported increased amygdala activation to negatively valenced stimuli in early blind participants [see Klinge et al. (2012)], suggesting that emotional vocalizations may be perceived as more salient when visual input is lacking. Furthermore, recent electrophysiological evidence suggests that early blindness relates to facilitated brain processing of authenticity of vocal emotional expressions (Sarzedas et al., 2024) and prior evidence showed ERP attention effects in blind individuals across a broader range of different emotional prosodies compared to sighted individuals (Topalidis et al., 2020).

To shed light on whether a lack of visual input and experience is associated with an improved capacity to interpret emotional vocalizations, we presented the same intense emotional vocalizations [as previously used by Atias et al. (2019)] to a group of early blind individuals and a group of normally sighted controls. Participants were asked to evaluate the stimuli as emotionally positive or negative, as well as to rate their intensity.

Mean valence ratings from both positive (M = −1.22, SD = 1.37) and negative vocalizations (M = −1.48, SD = 1.41) were found to be in the negative valence range. This suggests that overall, participants tended to interpret all vocalizations as emotionally, negative (see Figure 1A). The mixed repeated-measures ANOVA on valence recognition revealed a significant main effect of Valence, F(1,42) = 9.01, p = 0.005, η_p² = 0.18, indicating that participants identified positive vocalizations as slightly more positive than negative ones. Neither the main effect of Group, F(1,42) = 2.56, p = 0.12, nor the interaction Valence by Group, F(1,42) = 0.0007, p = 0.98, reached statistical significance.

Regarding intensity scores, vocalizations were overall perceived at an intermediary level of intensity (just above the middle score of the intensity 1–9 Likert scale) (see Figure 1B). The mixed repeated-measures ANOVA on intensity scores revealed a main effect of Valence, F(1,42) = 103.51, p < 0.001, η_p² = 0.71, indicating that negative vocalizations were rated more intense than positive. Furthermore, the factor of Group was significant, F(1,42) = 4.86, p = 0.033, η_p² = 0.10, indicating that blind participants perceived the vocalizations to be more intense (M = 5.94, SD = 1.05) than sighted participants (M = 4.95, SD = 1.14). The interaction Valence by Group, F(1,42) = 0.915, p = 0.34, was not statistically significant.

The capacity to understand others’ emotions and react accordingly is a key social ability that may be compromised in case of a profound sensory loss. Here we specifically investigated whether early blindness affects the capacity to interpret emotional vocalizations, whose valence may be difficult to recognize in the absence of a meaningful context (e.g., Atias et al., 2019). We found that blind and sighted participants found discriminating the valence of intense emotional vocalizations similarly difficult when no other contextual information was provided. Nonetheless, blind participants rated the affective vocalizations as more intense than sighted controls, possibly reflecting a higher perceived saliency.

The intensity of non-verbal vocalizations is characterized by specific acoustic profiles. Intense vocal stimuli are associated with increased duration, high fundamental frequency (pitch), and more high-frequency energy in the spectrum (bright timbre) (Anikin, 2020). These features inform about the significance and saliency of the stimuli, therefore attracting (automatic/bottom-up) attention and facilitating information prioritization and efficiency of appraisal (Belin and Zatorre, 2015; Anikin, 2020). For instance, screams have acoustic features dedicated to alarm signals but do not inform about the specific content meaning (fear vs. anger) of the emotion conveyed or the speaker’s identity (Anikin, 2020). Our findings indicate that blind participants are more sensitive than sighted individuals to these intensity-related acoustic characteristics, possibly because they represent a unique or primary source of sensory information during social interactions signaling danger or possible negative outcomes. Similar evidence was previously reported by Klinge et al. (2010, 2012) also showing that blindness is associated with greater amygdala sensitivity in response to emotional prosodic stimuli (and in particular negative ones) suggesting a neural mechanism for this observed behavioral compensation (Klinge et al., 2010, 2012). Interestingly, this finding may also be interpreted considering the higher interoceptive capacity observed in blind individuals (Radziun et al., 2023). High interoceptive ability is associated with increased sensitivity to the emotions of others (Terasawa et al., 2014) and with increased connectivity of the salience network (Chong et al., 2017). This may contribute to the perceived higher intensity of emotional stimuli by blind individuals, a hypothesis that deserves further investigation.

Our data show that blind participants performed similarly to the sighted controls in processing the valence of spontaneous intense vocal expressions. In particular, and regardless of visual status, participants rated both positive and negative vocalizations in the negative range of the valence scale, thus not reliably differentiating their valence, and replicating the previous findings of Atias et al. (2019). Hence, blindness does not seem to lead to an advantage in emotional recognition of vocal stimuli [see Gamond et al. (2017), Martins et al. (2019), Chen et al. (2022), and Sarzedas et al. (2024)], as with other domains such as spatial auditory localization or sentence comprehension (Battal et al., 2020; Loiotile et al., 2020). One possibility is that blind individuals do not develop better competencies in this domain because they may rely on other sources of contextual information such as verbal content [see Occelli et al. (2017)]. Indeed, in social interactions in everyday life, vocal emotional cues are usually accompanied by verbal semantic information. This may be more common for blind individuals who often rely on what other people say or on audio descriptions when accessing certain forms of media (e.g., television). Moreover, prior evidence also suggests that blindness may not affect social cognition abilities overall, which may develop via other sources [Ricciardi et al., 2009; for a recent meta-analysis, see Arioli et al. (2021)]. For instance, blind and sighted individuals similarly evaluate others’ trustworthiness based on the pitch of the voice (Oleszkiewicz et al., 2017) and similarly form social impressions of others [Ferrari et al., 2017; Arioli et al., 2024; see Bedny et al. (2009) and Arioli et al. (2021), for evidence on similar neural correlates of social cognition in the blind and sighted]. Overall, this suggests that blindness-related compensatory mechanisms in the auditory domain are restricted to specific auditory functions such as navigation/orientation, as well as spatial and language processing (Cattaneo et al., 2008, 2010b; Handjaras et al., 2016; Battal et al., 2020; Loiotile et al., 2020), rather than being generalized to the socioemotional domain. The advantage of blind individuals in language processing might be sufficient to ensure proper emotional and social competence or at least, socioemotional skills that are comparable to that of sighted people.

In interpreting our findings, some limitations should be acknowledged. The first limitation relates to the stimuli we used. Indeed, to make the task more sensitive to detect possible group differences, we intentionally employed ambiguous intense vocalizations (Atias et al., 2019) without providing any contextual information. The task may have been therefore too demanding hiding possible group differences. Prior evidence suggests that emotional stimuli may influence behavior only when their emotional content is relevant to the participants’ goal given the context in which they operate (e.g., Calbi et al., 2022; Mancini et al., 2022; Mirabella et al., 2023). Accordingly, future studies may test whether “contextualized” vocalizations elicit different reactions in blind and sighted participants, possibly also employing more nuanced vocalizations [see Cowen et al. (2019)] as well as physiological indexes of emotional arousal (e.g., skin conductance response, heart rate). A second limitation relates to the lack of control tasks in our study. In particular, one may object that the higher intensity rates reported by blind participants may represent an unspecific effect possibly extending to other auditory (non-emotional) stimuli or reflecting response biases, and this possibility should be controlled for in an additional task. However, in our study we specifically asked participants to evaluate the intensity of the emotion expressed and not “loudness” of the sounds, and prior studies have shown that early blind individuals exhibit similar sensitivity thresholds for (neutral) sounds’ intensity discrimination [Voss and Zatorre, 2012; for a review see Sabourin et al. (2022)].

In summary, although our findings are preliminary and deserve further investigation, our study represents the first attempt to explore the effect of early blindness on the perception of spontaneous emotional vocalizations.

The dataset for this article is not publicly available due to concerns regarding participants’ anonymity. Requests to access the datasets should be directed to the corresponding author.

The studies involving humans were approved by Research Integrity and Ethics Committee of the University of Bergamo. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

CF: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Visualization, Writing – original draft, Writing – review & editing. MA: Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing – original draft, Writing – review & editing. DA: Methodology, Writing – review & editing. LM: Supervision, Validation, Writing – review & editing. ZC: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.