Evolutionary loss of complexity in human vocal anatomy as an adaptation for speech

Language evolution in primates

Human speech evolution is not just about having a speech-ready brain and vocal apparatus.

Chimpanzee utterances refute purported missing links for novel vocalizations and syllabic speech

Nonhuman great apes have been claimed to be unable to learn human words due to a lack of the necessary neural circuitry. We recovered original footage of two enculturated chimpanzees uttering the word "mama" and subjected recordings to phonetic analysis. Our analyses demonstrate that chimpanzees are capable of syllabic production, achieving consonant-to-vowel phonetic contrasts via the simultaneous recruitment and coupling of voice, jaw and lips. In an online experiment, human listeners naive to the recordings' origins reliably perceived chimpanzee utterances as syllabic utterances, primarily as "ma-ma", among foil syllables. Our findings demonstrate that in the absence of direct data-driven examination, great ape vocal production capacities have been underestimated. Chimpanzees possess the neural building blocks necessary for speech.

The domestication of the larynx: The neural crest connection

Wolves howl and dogs bark, both are able to produce variants of either vocalization, but we see a distinct difference in usage between wild and domesticate. Other domesticates also show distinct changes to their vocal output: domestic cats retain meows, a distinctly subadult trait in wildcats. Such differences in acoustic output are well-known, but the causal mechanisms remain little-studied. Potential links between domestication and vocal output are intriguing for multiple reasons, and offer a unique opportunity to explore a prominent hypothesis in domestication research: the neural crest/domestication syndrome hypothesis. This hypothesis suggests that in the early stages of domestication, selection for tame individuals decreased neural crest cell (NCCs) proliferation and migration, which led to a downregulation of the sympathetic arousal system, and hence reduced fear and reactive aggression. NCCs are a transitory stem cell population crucial during embryonic development that tie to diverse tissue types and organ systems. One of these neural-crest derived systems is the larynx, the main vocal source in mammals. We argue that this connection between NCCs and the larynx provides a powerful test of the predictions of the neural crest/domestication syndrome hypothesis, discriminating its predictions from those of other current hypotheses concerning domestication.

What's special about human speech? A student exercise for comparing speech production between humans and chimpanzees

Modern humans and chimpanzees share a common ancestor on the phylogenetic tree, yet chimpanzees do not spontaneously produce speech or speech sounds. The lab exercise presented in this paper was developed for undergraduate students in a course entitled "What's Special About Human Speech?" The exercise is based on acoustic analyses of the words "cup" and "papa" as spoken by Viki, a home-raised, speech-trained chimpanzee, as well as the words spoken by a human. The analyses allow students to relate differences in articulation and vocal abilities between Viki and humans to the known anatomical differences in their vocal systems. Anatomical and articulation differences between humans and Viki include (1) potential tongue movements, (2) presence or absence of laryngeal air sacs, (3) presence or absence of vocal membranes, and (4) exhalation vs inhalation during production.

Spectrotemporal cues and attention jointly modulate fMRI network topology for sentence and melody perception

Speech and music are two fundamental modes of human communication. Lateralisation of key processes underlying their perception has been related both to the distinct sensitivity to low-level spectrotemporal acoustic features and to top-down attention. However, the interplay between bottom-up and top-down processes needs to be clarified. In the present study, we investigated the contribution of acoustics and attention to melodies or sentences to lateralisation in fMRI functional network topology. We used sung speech stimuli selectively filtered in temporal or spectral modulation domains with crossed and balanced verbal and melodic content. Perception of speech decreased with degradation of temporal information, whereas perception of melodies decreased with spectral degradation. Applying graph theoretical metrics on fMRI connectivity matrices, we found that local clustering, reflecting functional specialisation, linearly increased when spectral or temporal cues crucial for the task goal were incrementally degraded. These effects occurred in a bilateral fronto-temporo-parietal network for processing temporally degraded sentences and in right auditory regions for processing spectrally degraded melodies. In contrast, global topology remained stable across conditions. These findings suggest that lateralisation for speech and music partially depends on an interplay of acoustic cues and task goals under increased attentional demands.

Central Autonomic Mechanisms Involved in the Control of Laryngeal Activity and Vocalization

In humans, speech is a complex process that requires the coordinated involvement of various components of the phonatory system, which are monitored by the central nervous system. The larynx in particular plays a crucial role, as it enables the vocal folds to meet and converts the exhaled air from our lungs into audible sounds. Voice production requires precise and sustained exhalation, which generates an air pressure/flow that creates the pressure in the glottis required for voice production. Voluntary vocal production begins in the laryngeal motor cortex (LMC), a structure found in all mammals, although the specific location in the cortex varies in humans. The LMC interfaces with various structures of the central autonomic network associated with cardiorespiratory regulation to allow the perfect coordination between breathing and vocalization. The main subcortical structure involved in this relationship is the mesencephalic periaqueductal grey matter (PAG). The PAG is the perfect link to the autonomic pontomedullary structures such as the parabrachial complex (PBc), the Kölliker-Fuse nucleus (KF), the nucleus tractus solitarius (NTS), and the nucleus retroambiguus (nRA), which modulate cardiovascular autonomic function activity in the vasomotor centers and respiratory activity at the level of the generators of the laryngeal-respiratory motor patterns that are essential for vocalization. These cores of autonomic structures are not only involved in the generation and modulation of cardiorespiratory responses to various stressors but also help to shape the cardiorespiratory motor patterns that are important for vocal production. Clinical studies show increased activity in the central circuits responsible for vocalization in certain speech disorders, such as spasmodic dysphonia because of laryngeal dystonia.

Twin vocal folds as a novel evolutionary adaptation for vocal communications in lemurs

Primates have varied vocal repertoires to communicate with conspecifics and sometimes other species. The larynx has a central role in vocal source generation, where a pair of vocal folds vibrates to modify the air flow. Here, we show that Madagascan lemurs have a unique additional pair of folds in the vestibular region, parallel to the vocal folds. The additional fold has a rigid body of a vocal muscle branch and it is covered by a stratified squamous epithelium, equal to those of the vocal fold. Such anatomical features support the hypothesis that it also vibrates in a manner like the vibrations that occur in the vocal folds. To examine the acoustic function of the two pairs of folds, we made a silicone compound model to demonstrate that they can simultaneously vibrate to lower the fundamental frequency and increase vocal efficiency. Similar acoustic effects are achieved using different features of the larynx for the other primates, e.g., by vibrating multiple sets of ventricular folds in several species and further by an evolutionary modification of enlarged larynx in howler monkeys. Our multidisciplinary approaches found that these functions were acquired through a unique evolutionary adaptation of the twin vocal folds in Madagascan lemurs.

Development of the hyolaryngeal architecture in horseshoe bats: insights into the evolution of the pulse generation for laryngeal echolocation

BACKGROUND

The hyolaryngeal apparatus generates biosonar pulses in the laryngeally echolocating bats. The cartilage and muscles comprising the hyolarynx of laryngeally echolocating bats are morphologically modified compared to those of non-bat mammals, as represented by the hypertrophied intrinsic laryngeal muscle. Despite its crucial contribution to laryngeal echolocation, how the development of the hyolarynx in bats differs from that of other mammals is poorly documented. The genus Rhinolophus is one of the most sophisticated laryngeal echolocators, with the highest pulse frequency in bats. The present study provides the first detailed description of the three-dimensional anatomy and development of the skeleton, cartilage, muscle, and innervation patterns of the hyolaryngeal apparatus in two species of rhinolophid bats using micro-computed tomography images and serial tissue sections and compares them with those of laboratory mice. Furthermore, we measured the peak frequency of the echolocation pulse in active juvenile and adult individuals to correspond to echolocation pulses with hyolaryngeal morphology at each postnatal stage.

RESULTS

We found that the sagittal crests of the cricoid cartilage separated the dorsal cricoarytenoid muscle in horseshoe bats, indicating that this unique morphology may be required to reinforce the repeated closure movement of the glottis during biosonar pulse emission. We also found that the cricothyroid muscle is ventrally hypertrophied throughout ontogeny, and that the cranial laryngeal nerve has a novel branch supplying the hypertrophied region of this muscle. Our bioacoustic analyses revealed that the peak frequency shows negative allometry against skull growth, and that the volumetric growth of all laryngeal cartilages is correlated with the pulse peak frequency.

CONCLUSIONS

The unique patterns of muscle and innervation revealed in this study appear to have been obtained concomitantly with the acquisition of tracheal chambers in rhinolophids and hipposiderids, improving sound intensity during laryngeal echolocation. In addition, significant protrusion of the sagittal crest of the cricoid cartilage and the separated dorsal cricoarytenoid muscle may contribute to the sophisticated biosonar in this laryngeally echolocating lineage. Furthermore, our bioacoustic data suggested that the mineralization of these cartilages underpins the ontogeny of echolocation pulse generation. The results of the present study provide crucial insights into how the anatomy and development of the hyolaryngeal apparatus shape the acoustic diversity in bats.

Nonlinear dynamics and chaos in a vocal-ventricular fold system

In humans, ventricular folds are located superiorly to the vocal folds. Under special circumstances such as voice pathology or singing, they vibrate together with the vocal folds to contribute to the production of voice. In the present study, experimental data measured from physical models of the vocal and ventricular folds were analyzed in the light of nonlinear dynamics. The physical models provide a useful experimental framework to study the biomechanics of human vocalizations. Of particular interest in this experiment are co-oscillations of the vocal and ventricular folds, occasionally accompanied by irregular dynamics. We show that such a system can be regarded as two coupled oscillators, which give rise to various cooperative behaviors such as synchronized oscillations with a 1:1 or 1:2 frequency ratio and desynchronized oscillations with torus or chaos. The insight gained from the view of nonlinear dynamics should be of significant use for the diagnosis of voice pathologies, such as ventricular fold dysphonia.

Beyond speech: Exploring diversity in the human voice

Humans have evolved voluntary control over vocal production for speaking and singing, while preserving the phylogenetically older system of spontaneous nonverbal vocalizations such as laughs and screams. To test for systematic acoustic differences between these vocal domains, we analyzed a broad, cross-cultural corpus representing over 2 h of speech, singing, and nonverbal vocalizations. We show that, while speech is relatively low-pitched and tonal with mostly regular phonation, singing and especially nonverbal vocalizations vary enormously in pitch and often display harsh-sounding, irregular phonation owing to nonlinear phenomena. The evolution of complex supralaryngeal articulatory spectro-temporal modulation has been critical for speech, yet has not significantly constrained laryngeal source modulation. In contrast, articulation is very limited in nonverbal vocalizations, which predominantly contain minimally articulated open vowels and rapid temporal modulation in the roughness range. We infer that vocal source modulation works best for conveying affect, while vocal filter modulation mainly facilitates semantic communication.

Preliminary evidence for selective cortical responses to music in one-month-old infants

Prior studies have observed selective neural responses in the adult human auditory cortex to music and speech that cannot be explained by the differing lower-level acoustic properties of these stimuli. Does infant cortex exhibit similarly selective responses to music and speech shortly after birth? To answer this question, we attempted to collect functional magnetic resonance imaging (fMRI) data from 45 sleeping infants (2.0- to 11.9-weeks-old) while they listened to monophonic instrumental lullabies and infant-directed speech produced by a mother. To match acoustic variation between music and speech sounds we (1) recorded music from instruments that had a similar spectral range as female infant-directed speech, (2) used a novel excitation-matching algorithm to match the cochleagrams of music and speech stimuli, and (3) synthesized "model-matched" stimuli that were matched in spectrotemporal modulation statistics to (yet perceptually distinct from) music or speech. Of the 36 infants we collected usable data from, 19 had significant activations to sounds overall compared to scanner noise. From these infants, we observed a set of voxels in non-primary auditory cortex (NPAC) but not in Heschl's Gyrus that responded significantly more to music than to each of the other three stimulus types (but not significantly more strongly than to the background scanner noise). In contrast, our planned analyses did not reveal voxels in NPAC that responded more to speech than to model-matched speech, although other unplanned analyses did. These preliminary findings suggest that music selectivity arises within the first month of life. A video abstract of this article can be viewed at https://youtu.be/c8IGFvzxudk. RESEARCH HIGHLIGHTS: Responses to music, speech, and control sounds matched for the spectrotemporal modulation-statistics of each sound were measured from 2- to 11-week-old sleeping infants using fMRI. Auditory cortex was significantly activated by these stimuli in 19 out of 36 sleeping infants. Selective responses to music compared to the three other stimulus classes were found in non-primary auditory cortex but not in nearby Heschl's Gyrus. Selective responses to speech were not observed in planned analyses but were observed in unplanned, exploratory analyses.

Singing more, singing harsher: occurrence of nonlinear phenomena in a primate' song

Nonlinear phenomena (NLP) in animal vocalizations arise from irregularities in the oscillation of the vocal folds. Various non-mutually exclusive hypotheses have been put forward to explain the occurrence of NLP, from adaptive to physiological ones. Non-human primates often display NLP in their vocalizations, yet the communicative role of these features, if any, is still unclear. We here investigate the occurrence of NLP in the song of a singing primate, the indri (Indri indri), testing for the effect of sex, age, season, and duration of the vocal display on their emission. Our results show that NLP occurrence in indri depends on phonation, i.e., the cumulative duration of all the units emitted by an individual, and that NLP have higher probability to be emitted in the later stages of the song, probably due to the fatigue indris may experience while singing. Furthermore, NLP happen earlier in the vocal display of adult females than in that of the adult males, and this is probably due to the fact that fatigue occurs earlier in the former because of a greater contribution within the song. Our findings suggest, therefore, that indris may be subjected to physiological constraints during the singing process which may impair the production of harmonic sounds. However, indris may still benefit from emitting NLP by strengthening the loudness of their signals for better advertising their presence to the neighboring conspecific groups.

Comparative physiological anthropogeny: exploring molecular underpinnings of distinctly human phenotypes

Anthropogeny is a classic term encompassing transdisciplinary investigations of the origins of the human species. Comparative anthropogeny is a systematic comparison of humans and other living nonhuman hominids (so-called "great apes"), aiming to identify distinctly human features in health and disease, with the overall goal of explaining human origins. We begin with a historical perspective, briefly describing how the field progressed from the earliest evolutionary insights to the current emphasis on in-depth molecular and genomic investigations of "human-specific" biology and an increased appreciation for cultural impacts on human biology. While many such genetic differences between humans and other hominids have been revealed over the last two decades, this information remains insufficient to explain the most distinctive phenotypic traits distinguishing humans from other living hominids. Here we undertake a complementary approach of "comparative physiological anthropogeny," along the lines of the preclinical medical curriculum, i.e., beginning with anatomy and considering each physiological system and in each case considering genetic and molecular components that are relevant. What is ultimately needed is a systematic comparative approach at all levels from molecular to physiological to sociocultural, building networks of related information, drawing inferences, and generating testable hypotheses. The concluding section will touch on distinctive considerations in the study of human evolution, including the importance of gene-culture interactions.

Life of p: A consonant older than speech

Which sounds composed the first spoken languages? Archetypal sounds are not phylogenetically or archeologically recoverable, but comparative linguistics and primatology provide an alternative approach. Labial articulations are the most common speech sound, being virtually universal across the world's languages. Of all labials, the plosive 'p' sound, as in 'Pablo Picasso', transcribed /p/, is the most predominant voiceless sound globally and one of the first sounds to emerge in human infant canonical babbling. Global omnipresence and ontogenetic precocity imply that /p/-like sounds could predate the first major linguistic diversification event(s) in humans. Indeed, great ape vocal data support this view, namely, the only cultural sound shared across all great ape genera is articulatorily homologous to a rolling or trilled /p/, the 'raspberry'. /p/-like labial sounds represent an 'articulatory attractor' among living hominids and are likely among the oldest phonological features to have ever emerged in linguistic systems.

Large Language Models and the Reverse Turing Test

Large language models (LLMs) have been transformative. They are pretrained foundational models that are self-supervised and can be adapted with fine-tuning to a wide range of natural language tasks, each of which previously would have required a separate network model. This is one step closer to the extraordinary versatility of human language. GPT-3 and, more recently, LaMDA, both of them LLMs, can carry on dialogs with humans on many topics after minimal priming with a few examples. However, there has been a wide range of reactions and debate on whether these LLMs understand what they are saying or exhibit signs of intelligence. This high variance is exhibited in three interviews with LLMs reaching wildly different conclusions. A new possibility was uncovered that could explain this divergence. What appears to be intelligence in LLMs may in fact be a mirror that reflects the intelligence of the interviewer, a remarkable twist that could be considered a reverse Turing test. If so, then by studying interviews, we may be learning more about the intelligence and beliefs of the interviewer than the intelligence of the LLMs. As LLMs become more capable, they may transform the way we interact with machines and how they interact with each other. Increasingly, LLMs are being coupled with sensorimotor devices. LLMs can talk the talk, but can they walk the walk? A road map for achieving artificial general autonomy is outlined with seven major improvements inspired by brain systems and how LLMs could in turn be used to uncover new insights into brain function.

What made us "hunter-gatherers of words"

This paper makes three interconnected claims: (i) the "human condition" cannot be captured by evolutionary narratives that reduce it to a recent 'cognitive modernity', nor by narratives that eliminates all cognitive differences between us and out closest extinct relatives, (ii) signals from paleogenomics, especially coming from deserts of introgression but also from signatures of positive selection, point to the importance of mutations that impact neurodevelopment, plausibly leading to temperamental differences, which may impact cultural evolutionary trajectories in specific ways, and (iii) these trajectories are expected to affect the language phenotypes, modifying what is being learned and how it is put to use. In particular, I hypothesize that these different trajectories influence the development of symbolic systems, the flexible ways in which symbols combine, and the size and configurations of the communities in which these systems are put to use.

Physical modeling of the vocal membranes and their influence on animal voice production

The vocal membrane, i.e., an extended part of the vocal fold, is observed in a wide range of species including bats and primates. A theoretical study [Mergell, Fitch, and Herzel (1999). J. Acoust. Soc. Am. 105(3), 2020-2028] predicted that the vocal membranes can make the animal vocalizations more efficient by lowering the phonation threshold pressure. To examine this prediction, a synthetic model of the vocal membrane was developed, and its oscillation properties were examined. The experiments revealed that the phonation threshold pressure was lower in the vocal membrane model compared to that in a model with no vocal membrane. Chaotic oscillations were observed as well.

When less is more in the evolution of language

Did loss of vocal fold membranes typical of nonhuman primates enable human speech?