Methods
The vocalizations of 4 female and 2 male monkeys were categorized according to their behavioural context as pleading, content, commanding, frightened, angry and emotionally neutral. Twenty women simulated similar connotations using the name "Saara" (Finnish) or "Sarah" (British English). The samples agreed upon by more than 30% of Finnish or British English listeners were included in the comparison. The signals were sampled at 22 kHz using 16-bit quantization. The monkey vocalizations and speech samples (the long vowel segments) were compared with respect to variation of duration, fundamental frequency (F0), singnal waveform and patterning of spectral energy distribution.

Results
Monkey vocalizations were of 0.3-1.5 s and word samples of 0.5-1 s duration. F0 was 500-5000 Hz in monkeys and 100-500 Hz in humans. The monkey vocalizations differed from the speech samples with respect to variation of F0 peak position. There were many similarities in short-time features determinant for voice quality.
In both monkey and man, "neutral", "pleading" and "commanding" were not distinguished by mean F0 but the signal waveforms and spectral patterns were clearly different. "Anger" and "fear" were expressed with a higher mean F0 than all other connotations and the highest means were measured for "fear". "Anger" and "fear" differed with respect to spectral energy distribution. Regular low-frequency (< 100 Hz) modulations or abrupt changes in amplitude were encountered in the aggressive vocalizations of both monkey and man. The content grunts of monkeys were soft noisy sounds with 80-90 Hz periodicity. Grunting sounds do not apear in Finnish or English speech. The few imitations of "content" agreed upon by the listeners suggested a lax vocal fold closure and breathy (soft and noisy) voice quality.

Conclusions
The comparisons suggest that monkeys and man share acoustic cues and vocal gestures for anger, fear, command/dominance, plea/submission and emotional neutrality. The shared acoustic patterns derive from homologies in the mechanisms of vocalization and its neural control. Man and macaques may also share acoustic cues, although not the production mechanisms, for contentment.
The acoustic features differentiating among emotional connotations are known to discriminate among sounds with high and low perceived tonality, pleasantness, roughness, sharpness and loudness, for instance. Interprimate comparisons may aid the recognition of communicative cues with high developmental significance postnatally.