Linking human male vocal parameters to perceptions, body morphology, strength and hormonal profiles in contexts of sexual selection

Sexual selection has significantly shaped acoustic signals across diverse species, including humans. Deep, resonant voices in men are hypothesized to play a role in attracting mates and intimidating rivals. However, much of the existing evidence comes from studies with low ecological validity, manipulating single vocal parameters artificially. This study aimed to address these limitations by examining the relationship between naturally occurring vocal parameters and various measures of male quality in a large sample size. The study explores the adaptive significance of attending to specific vocal cues in both intrasexual (male-male) and intersexual (male-female) contexts of sexual selection, testing the hypothesis that these parameters reflect condition and formidability, providing reliable information about the speaker's quality as a potential mate or rival. This understanding is crucial for comprehending the evolutionary pressures that have shaped human vocal communication and its role in reproductive success.

Previous research has established a link between acoustic parameters, particularly fundamental frequency (f₀) and formant frequencies, and perceptions of attractiveness, dominance, and masculinity in men. Lower f₀ is generally associated with higher attractiveness and dominance ratings, potentially linked to greater mating and reproductive success. However, the underlying reasons for this association remain unclear. Costly signaling theory suggests that honest signals of quality are maintained through various costs, but the costs associated with men's f₀ are debated. Some studies point to weak receiver-independent costs, questioning the honesty of this signal. Others argue that f₀ is partly honest, correlating with testosterone and indicators of physical condition, notably having a stronger negative correlation with testosterone among men with lower cortisol levels. Formant frequencies, related to vocal tract length, show weaker, albeit some, correlations with body size but equivocal relationships with strength. Existing studies primarily focus on individual parameter manipulation, linear relationships, and lack ecological validity using artificial stimuli.

This preregistered study (https://osf.io/nrmpf/) involved 160 heterosexual men (mean age 24.28 years) recruited from the University of Göttingen. Participants provided standardized 5-second video recordings of themselves describing positive attributes in three chosen domains. Saliva samples were collected to measure baseline cortisol and testosterone levels. Body morphology was assessed using 3D body scans, extracting numerous dimensions including body volume, girth measurements, and ratios. Strength was measured using handgrip and upper-body strength assessments. A separate group of 240 raters (120 men, 120 women, mean age ≈20 years) from Pennsylvania State University rated the voice recordings for short-term and long-term attractiveness, and social and physical dominance. Voice recordings were analyzed using PRAAT software to extract mean f₀, formant frequencies (F1-F4), and measures of jitter and shimmer. A composite measure of formant structure (Pf) was calculated as the average standardized formant value for the first four formants. Statistical analyses, including multiple regressions, curvilinear analyses, and meta-analyses, were employed to test pre-registered and exploratory hypotheses.

The study found support for many hypotheses. Lower f₀ negatively predicted both short-term and long-term attractiveness in a negatively curvilinear relationship, suggesting a trade-off between potential benefits (genetic quality) and costs (e.g., reduced investment, risk of violence) associated with highly masculine voices. F₀ was a stronger predictor of short-term attractiveness, supporting the prioritization of genetic fitness in short-term mating contexts. Formant position (Pf) had weaker effects on attractiveness ratings, negatively correlating with long-term attractiveness in a curvilinear fashion. F₀ negatively predicted physical dominance but not social dominance. Pf, however, negatively predicted both physical and social dominance. In multiple regression analyses, f₀ and Pf independently predicted physical dominance ratings. Testosterone and cortisol interacted in predicting f₀, with higher testosterone negatively linked to f₀ particularly when cortisol levels were low. This interaction was confirmed by a meta-analysis combining the current study's data with previous findings. Pf negatively correlated with strength and several body morphology measures. A mediation analysis showed that Pf, independently of f₀, mediated the relationship between composite body size and physical dominance. A meta-analysis revealed that f₀ mediates the relationship between height and physical dominance. Shimmer negatively correlated with social and physical dominance ratings, along with lower cortisol, whereas jitter showed no significant relationship with any dependent variables.

The findings support the hypothesis that male vocal parameters provide cues about condition and formidability relevant to sexual selection. F₀ emerged as a robust signal of attractiveness, potentially reflecting a trade-off between genetic benefits and potential costs. The interaction of testosterone and cortisol in predicting f₀ highlights the complex hormonal interplay influencing vocal characteristics. Pf and body size showed a consistent relationship, reflecting a link between vocal cues and physical formidability. The difference in the predictive power of f₀ and Pf on different types of dominance ratings suggests that these acoustic parameters convey distinct information regarding the speaker's qualities. The significant role of shimmer suggests that vocal quality also plays a role in assessing potential mates or rivals. This study expands our understanding of the evolutionary basis of human voice perception and the adaptive nature of attention to specific vocal cues in mate choice and rivalry.

This study demonstrates a significant link between male vocal parameters, perceptions of attractiveness and dominance, and underlying biological factors. F₀ and Pf, along with shimmer, function as signals reflecting underlying condition, affecting both intersexual and intrasexual selection. Future research should further investigate the receiver-dependent costs associated with these vocal cues, explore the role of jitter and shimmer in greater detail, and examine cross-cultural variation in these effects.

This study did not explore receiver-dependent costs associated with low f₀, limiting the comprehensive assessment of signal honesty. The median split of cortisol may have reduced statistical power in some analyses. The use of a standardized self-description task for recording vocal samples, while aiming for higher ecological validity than typical methods, might still limit generalizability compared to completely naturalistic conversation samples.