Mbem is a specialized genre of Igbo oral poetry performed within masquerade traditions. It is a hybrid of song and speech, ordinarily performed by a single masked chanter without musical instruments, relying on the chant itself for rhythmic structure. Unlike masquerade songs, which are sung by groups, mbem is chanted in a distinctive cadence between ordinary speech and singing, and its content ranges from eulogy and epithet to social commentary, historical events, and reflections on human existence. Because the masked speaker is usually unidentifiable, the form can be used to admonish and to cajole, and the literature links the chanted delivery to the sense of spiritual authority and ancestral presence attributed to the masks.

The study focused on two well-known chants associated with two respected practitioners. Odogwuanyammee, whose name translates as “the red-eyed brave”, is an Arondizuogu mask known for daring exploits, whose chants are accompanied by a troupe of performers and rhythmic bass tom-toms. Okwuanyịọnụ, “the mouth that will not tire of speech”, is a masquerade from Agulu that performs entirely unaccompanied, in a mid-tempo, sing-song style between melodic singing and measured speech, conveying social commentary, satire, advice, and self-praise. The chants take their names from the personae that perform them, and one of the analyzed chants invokes the Igbo proverb that gives the paper its title: Ikwi ikwi anaghị ebe akwa na nkiti, “the owl never cries in vain”, a saying that treats unusual occurrences as meaningful signs demanding interpretation.

Two male chanters, aged 36 and 31, one from Arondizuogu in Imo State and one from Agulu in Anambra State, recorded their chants and then recited the same texts in normal speech mode. Each chant was produced twice in each mode. Recordings were made with a Zoom Q8 camera and a Shure SM10 head-mounted microphone at a 48 kHz sampling rate, and tones and vowels were manually annotated in Praat. The analysis measured seven acoustic parameters: fundamental frequency, intensity, cepstral peak prominence, energy below 500 Hz, the Hammarberg index, and the first and second formants. Linear mixed-effects models with mode as a fixed effect were used to compare the three vocalization modes, with post hoc pairwise comparisons adjusted for multiple testing.

The most striking result concerns pitch. Fundamental frequency was consistently higher in the chant modes, roughly 220 to 300 Hz, than in speech mode, roughly 130 to 150 Hz, across all six lines of the analyzed material. The difference between speech and chant was most pronounced at the midpoint of each tone, with a model coefficient of −104.826 Hz for speech relative to the Odogwuanyammee chant (p < 0.001), and significant differences also appeared at the 20% and 80% measurement points. Importantly, the chant modes differed from speech primarily in overall pitch register rather than in the shape of the pitch contour, and the gradual terminal descent within each line was consistent with the tone-terracing described for the Igbo language.

That last point matters because Igbo is a register tone language in which pitch carries lexical meaning. The language has three level tones, high, low, and downstepped high, and minimal sets such as òké “rat” versus ókè “boundary” show how tone distinguishes words. The analysis found that the three tone categories maintained distinct fundamental frequency patterns across all modes, even though the magnitude of the distinctions varied, and the high tone consistently showed the highest values although that difference was not statistically significant once random effects were accounted for. In other words, chanters raised the entire pitch register substantially while preserving the tonal contrasts that keep the words intelligible, a balance the author characterizes as the maintenance of linguistic functionality within ritual performance.

Vowel formants also shifted across modes. For the first formant, Okwuanyịọnụ showed higher values than Odogwuanyammee (β = 81.51, p < 0.001) and speech showed significantly lower values (β = −108.18, p < 0.001), a pattern suggesting more open vocal-tract configurations in the chanted forms. For the second formant, Okwuanyịọnụ differed significantly from Odogwuanyammee (β = −251.633, p < 0.001), indicating a more retracted tongue position, while speech showed no significant difference from Odogwuanyammee. Despite these stylistic modifications, the relative relationships among vowels were preserved, maintaining the phonological contrasts of the Igbo vowel system, which is organized into two harmonic sets by advanced tongue root.

Intensity and voice-quality measures completed the picture. Odogwuanyammee showed higher vocal intensity than both speech mode, which was substantially quieter (β = −23.83 dB, p < 0.001), and Okwuanyịọnụ, which was modestly quieter (β = −5.56 dB, p < 0.001). Speech concentrated more energy in the band below 500 Hz than Odogwuanyammee, while Okwuanyịọnụ showed a slight but significant decrease in that band. Cepstral peak prominence, which indexes the clarity of the vocal signal, was substantially lower in speech than in Odogwuanyammee (β = −13.80, p < 0.001), meaning the chant modes showed a more strongly voiced, projected quality rather than a breathy one, and the two chant modes did not differ significantly from each other on this measure. The Hammarberg index, which captures spectral balance, distinguished Okwuanyịọnụ from Odogwuanyammee (β = 5.12, p < 0.001) but not speech from Odogwuanyammee. Across vowels, these modifications were applied consistently within each mode.

Taken together, the findings indicate that the voice quality practitioners call “guttural” and “spiritual” is not produced by any single acoustic feature but by the coordinated manipulation of multiple parameters. The author notes that the perceptual label “guttural” is not straightforwardly reducible to the measured values: since the chant voice is actually higher-pitched and more modally voiced than speech, “guttural” here appears to denote the marked, effortful, projected character of the voice rather than low pitch or roughness in a literal sense. The data speak to the acoustic modifications themselves rather than to spiritual possession as such, and the author is careful to distinguish the measured observations from the cultural interpretations that scholars have attached to them.

The study also situates mbem within a broader comparative context. Whereas Yorùbá performers of genres such as Ìjálá rely heavily on vibrato and dynamic pitch modification, mbem achieves its marked voice through consistent modification of spectral balance and intensity, and unlike the more uniform intensity reported for Yorùbá genres, mbem shows significant variation between its two chant modes. The limitations are acknowledged explicitly: with only two chanters, one per mask persona, speaker-specific and genre-specific effects cannot be fully separated, each chant was recorded with only two repetitions per mode, and fundamental frequency and intensity were analyzed without speaker normalization. The findings are therefore presented as a preliminary study rather than population-level generalizations. Future work, the author suggests, could examine regional variation in mbem, the perceptual significance of these acoustic modifications for audiences, and comparative studies with other African tone languages. Returning to the proverb in the title, the analysis suggests a sense in which the owl never cries in vain: the chanting voice is not merely louder or higher speech but is systematically shaped along several acoustic dimensions, so that it is produced to be heard as marked and significant while remaining intelligible.

Subject of Research: Acoustic analysis of vocal expressions in Igbo mbem masquerade chants compared with normal speech

Article Title: “The owl never cries in vain”: an acoustic study of vocal expressions in mbem chants

Article References: Nwosu, V. (2026). “The owl never cries in vain”: an acoustic study of vocal expressions in mbem chants. Frontiers in Communication, 11, Article 1605728. https://doi.org/10.3389/fcomm.2026.1605728

Image Credits: AI Generated

DOI: 10.3389/fcomm.2026.1605728

Keywords: mbem chants, Igbo, masquerade, acoustic analysis, fundamental frequency, formants, voice quality, tone language, oral poetry, cepstral peak prominence, Hammarberg index, Nigeria