Page 31 - Fall 2011

cle mass in the arms (Puts, 2010). To test that idea, researchers estimated male “threat potential” by measuring height, bicep size, hand strength, salivary testosterone level, and inherent aggressiveness. Outcomes showed stronger cor- relations between F0 and formants, and size, strength, and testosterone level than previously reported for either height or weight (Puts et al., 2011; see also Sell et al., 2010). The resulting argument is that the male voice does provide important, reliable cues to vocalizer competitive capabilities, and that listeners are responding reasonably to those cues. Mate-choice competition Recent studies have also addressed the related question of whether vocal characteristics play an important role in mate-choice competition—here expecting both sexes to show such effects. The basic approach has been to ask lis- teners to rate the relative attractiveness of a variety of male and female voices, with pitch and resonance again being the critical variables. Testing females, it is common to find a preference for masculinized voices—meaning those with lower vocal pitch and resonances (Feinberg, 2008; Jones et al., 2010; Pisanski and Rendall, 2011). From a mate-choice perspective, these characteristics may represent hormone- related ornamentation that has emerged precisely due to being attractive to females (Feinberg, 2008). Although the relationship is modest, male salivary testosterone levels have indeed been found to be inversely correlated with both F0 and formant frequencies (Dabbs and Mallinger, 1999; Bruckert et al., 2006; Evans et al., 2008). Other evidence sup- porting an influence of mate-choice selection includes a sta- tistical correlation between these vocal characteristics and both number of children fathered (Apicella et al., 2007) and number of sexual encounters reported (Hodges-Simeon et al., 2011). Finally, listeners are more likely to expect infideli- ty from males with masculinized voices (O’Connor, 2011), which are also preferred more by females when approaching ovulation than at non-fertile times within the menstrual cycle (Feinberg et al., 2006). Males show approximately converse preferences, as might be expected. For instance, many studies have revealed a preference for higher-pitched female voices (Apicella and Feinberg, 2009; Pisanski and Rendall, 2011). This effect may be traceable to pitch as a fertility cue, with females being most fertile and having highest speaking F0 values in early adulthood (Stathopoulos et al., 2011). Males have also been found to prefer voices of females recorded when close to ovulation (Pipitone and Gallup, 2008), a point in the men- strual cycle that is also associated with increased vocal pitch (Bryant and Haselton, 2009)—although not uniquely so (Fischer et al., 2011). Other evidence includes increased pitch among females when believing they are communicat- ing with more masculinized and attractive males (Fraccaro et al., 2011), and women with higher-pitched voices are deemed more likely to exhibit infidelity (O’Connor, 2011). There are again complications, of course, but perhaps fewer than for intrasex competition in voice. For example, not all studies have found more masculine or feminine voices to be the most attractive. In at least one case, mid-range or aver- age voices in the opposite sex have been the most attractive for both male and female listeners (Hughes et al., 2010). The same work reported that all participants tended to speak at lower pitches when interacting with an attractive partner. While consistent with previously reported female prefer- ences, the result is inconsistent with other findings for males. There is also disagreement as to whether F0 and for- mants work separately (Hodges-Simeon et al., 2010; Jones et al., 2010) or synergistically (Feinberg et al., 2008; Feinberg et al., 2011). Conclusions This review has moved quickly and lightly over a variety of topics, each of which deserves much more thorough treat- ment. Nonetheless, the evidence covered underscores the fact that the human voice does have a long evolutionary his- tory and has been importantly shaped through shared phy- logeny with other species. Vocal-fold action and vocal-tract resonance have emerged as recurring themes, equally appli- cable to vocal production in humans and nonhuman mam- mals and creating substantive evolutionary connections between the two. However, it has also become apparent that hominin evolution also brought important changes. Understanding those changes raises questions that compar- isons to other primates and mammals alone may not fully address. Yet, combining clues from other species with evi- dence of novel human vocal characteristics may ultimately prove to be an effective means of shedding further light on hominin evolution overall. Three issues will be briefly fol- lowed up in closing, including the evident weakness of cor- relations between human vocal characteristics and physical features such as overall body-size, possible changes in vocal- fold stability over hominin evolution, and the intertwining of indexical and phonetic cuing in speech—the most unique of human vocalizations. Sexual selection and the voice—Understanding the role of sexual selection on the voice is a recent undertaking, and progress has been rapid. The overall approach has been vali- dated not only by finding evidence of sexual-selected vocal effects in both sexes, but also by the fact that outcomes are predictably somewhat different in males versus females. However, it is also difficult to avoid the feeling that an impor- tant piece of the puzzle is still missing. Correlations between vocal and physical characteristics are too weak, for example, and it is not satisfying to invoke global correlations from the world at large to explain an apparently illusory relationship between pitch and size. Another possibility is that those human mating decisions have become sufficiently complex over evolutionary time that vocal characteristics have lost an important link to physical characteristics that they once had. However, a more compelling explanation may emerge though more substantive recasting of key vocalizer traits as a combination of physical and psychological characteristics, such as threat potential. Overall, understanding sexual selec- tion effects in the human voice has some surprising, but interesting, complexities that may require imagination and re-thinking to untangle. 30 Acoustics Today, October 2011