Neil Todd. American Scientist. Volume 103, Issue 5. Sep/Oct 2015.
Most of us think of our senses as a group of five- vision, hearing, touch, smell, and taste-and can readily identify the parts of the body that let us process information about the world through each of these senses. Less easy to locate are the organs that underlie our sense of balance, although we depend on them at every moment. Hidden deep inside the ear, the organs of the vestibular system generally perform their role without requiring conscious attention. Yet an illness or injury that affects vestibular function can be devastating, as anyone with an innerear infection will testify. Sufferers may be almost entirely incapacitated, unable even to stand up, raise their head, or keep their eyes steady, without vomiting; as if this weren’t enough, they often feel acute anxiety as well. The unusually broad range of symptoms that can erupt when our sense of balance is disturbed, affecting everything from vision and hearing to posture, gut functioning, emotions, and cognition, should tell us just how crucial the vestibular system is for maintaining a normal physiological state.
In mammals the vestibular apparatus shares close quarters with fire cochlea, the main organ of hearing, in the inner-ear structure known as the labyrinth. The two systems also have other features in common, such as the use of hair-cells as mechanoreceptors, to perceive sounds and head movements and encode them in the form of nerve signals. Even the sensory nerves which run from the cochlea and the vestibular apparatus are bundled together in their passage up to the brain. This close partnership bears witness to a shared history of sensing vibration and gravity that can be traced back in geological time at least 500 million years.
From an evolutionary perspective, the cochlea is a new kid in the neighborhood. The earliest vertebrates, including fish, do not possess a cochlea or any organ remotely analogous to it. One could conclude from this that fish must be deaf. How can they hear without a cochlea? Attempts to explain this seeming paradox began nearly 100 years ago with the publication of a paper by Canadian physiologist John Tait, under the provocative title “Is All Hearing Cochlear?” After concerted efforts, researchers agreed that the answer to Tait’s title question was “no”: Fish, and also amphibians, use a system of noncochlear hearing, mediated by a set of organs containing otoliths, or “ear stones.” In fact, such mechanisms are widespread throughout the animal kingdom (although by convention they are termed statoliths in invertebrates). These small stones-actually crystals of calcium carbonate-are suspended in the inner ear and coupled to mechanoreceptors, which detect movement and transmit that sensory information in the form of neural activity. Thanks to this simple mechanism, fish can hear quite well, as any frustrated angler can attest. Moreover, some species of fish can be highly vocal, and this faculty plays an important biological role-not least for vocal courtship behavior, as will be discussed further in filis article.
By contrast, in the human ear, the cochlea is responsible for most of our perception and processing of sounds. Nevertheless, our ears contain otolith organs as well: a saccule, oriented in the vertical plane, and a utricle, oriented in the horizontal plane. The conventional view of the otolith organs in humans has long been that they simply serve the vestibular system-but could humans have retained an otolithic sense of hearing that augments our cochleabased auditory system? Certainly the existence of other parallel, more primitive sensory pathways has been documented before, as with the vomeronasal system, which makes use of pheromones and serves as a sort of auxiliary system to our conscious sense of smell. To explore this possibility, we must first examine the origins of acoustic sensing in the simplest life forms.
Starting Out with a Sense of Gravity
It is no accident that the two senses, hearing and balance, share an evolutionary history and that they are anatomically and physiologically related; this close relationship follows naturally from the laws of physics governing the detection of sound and motion. The world of invertebrates, and indeed of plants, offers many examples of such links.
Hearing and balance both arise from a fundamental requirement for life, which is knowing up from down. Even the earliest forms of life on Earth required this information both to navigate in their external environment and to control their internal environment in spite of changes in orientation. One simple way to do this would be by means of a light sensor and a single rule, that the direction of light is always “up.” Such a visual mechanism works well when there is plenty of direct and reflected light, but it has its limitations: at night, for instance, or in an environment such as the deep ocean, where the light cannot penetrate.
A more versatile solution, both for gaining knowledge of space and for distinguishing self-motion from object motion, is to make use of gravity, for which purpose the life form requires a sensor for detecting gravity. The simplest way to construct such a graviceptor is to make use of a difference in density between small, dense stones–the statoliths, which correspond to the otoliths in vertebrates–and a surrounding substance of lower density. The life form can then work out the direction of gravity, because no matter what its position, the little stones will always fall to the lowest point. Such a system has evolved independently many times, in both animals and plants. All major metazoan phyla have a statolithic sense mechanism-indeed, this observation has been used to argue that “up” versus “down” was the first sense to appear in the animal kingdom.
Many species of plants possess statocytes (cells containing one or more statoliths) in their growth tips that allow them to direct roots downward into the earth. A multicellular, gravitysensing organ that contains a statolith is referred to as a statocyst, and many examples of such an organ are found among invertebrates-for example, within the marine animals known as echinoderms, which have a remarkable ability to right themselves when turned upside down.
If an animal possesses a light sensor as well as a graviceptor, it is even better equipped to navigate in space. However, it must have some way of coordinating the two kinds of data so that the graviceptor and the light sensor do not provide contradictory information. One simple way to manage this is by having the two sensors mechanically coupled. Such close coupling evolved quite early in evolutionary terms, perhaps as much as 700 million years ago, in the cnidarians (a group including jelly fish, which possess stinging cells). The box jelly fish shows this coupling in its possession of light-sensing patches, image-forming eyes, and statoliths, closely and rigidly situated within a common sensory pod that hangs freely on a stalk within its bell-shaped body. Remarkably, these animals have four such pods, each containing a number of image-forming and non-image-forming eyes. As a result, whatever the orientation of the bell, the pod always aligns with gravity and the eyes always maintain the same vertical alignment. Contrary to our popular conception of jelly fish, these species are able to navigate with this coordinated sensory mechanism to find food.
In more highly evolved invertebrates, for example in the octopus, we find a system that can detect both rotational and linear acceleration, as well as orientation. The pair of statocysts in the octopus shows a close parallel to the vertebrate vestibular system-and allows the octopus both to maintain its orientation in space and to stabilize retinal images for complex visual processing via an equivalent of the vertebrate vestibular-ocular reflexes. Of particular interest is the fact that the octopus statocyst also responds to vibration as well as to gravity and acceleration. Any sound source which causes local fluid vibrations (the acoustic near field) will result in the vibration being transmitted to the animal’s body, hence causing relative motion of the statolith contained within its body. Thus we see that an organ originally evolved to detect gravity becomes an organ to perceive sound. The many cases of parallel evolution of statolith systems in invertebrates, and the easy conversion of a statolith organ from linear acceleration sensor to audio frequency vibration sensor, collectively suggest the likely scenario for the emergence of an otolithic hearing mechanism in vertebrates that appeared in the geological record around the time of the Cambrian explosion, more than 540 million years ago.
Beyond the Basics
All fish use at least one of their otolith organs for hearing as well as balance, but in addition some species have developed ancillary structures that enable them to improve on the basic nearfield hearing of the octopus. In goldfish (Carassius auratus), for example, the swim bladder may serve to convert waves of sound pressure (the acoustic far field) to local displacement. Goldfish and other species with this adaptation also possess a specialized structure that couples the swim bladder directly to the inner ear containing the otolith organs. With the addition of such structures, fish hearing can be remarkably acute, so that in the low-frequency range (between about 100 and 300 hertz) the hearing sensitivity of the goldfish approaches that of the human ear.
It’s obviously useful for an animal to be able to navigate in space so as to avoid predation and to find food for itself. Having the ability to perceive sound and vibration gives the animal extra advantages in being able to detect prey or avoid predation by building a kind of acoustic picture of the world around it.
Another advantage that hearing might confer, less obvious but equally important, is the possibility of using it to find a mate. The males of many species of fish produce sounds as an element in courtship, using their swim bladder as a kind of talking drum. They have special drumming muscles attached to the swim bladder for this purpose. The swim bladder itself, interestingly, is innervated in some species by the hypoglossal nerve (the same nerve that in humans controls the tongue), and in others by the vagus nerve, which in humans innervates fire larynx.
A classic example of a vocal fish can be found in the case of the haddock (Mellanogrammus aeglefinus), the species often cooked for a meal of fish and chips. The vocalizing male usually starts out trying to attract a female by gentle drumming, but as courtship proceeds the drumming frequency increases to the extent that it sounds like a kind of buzzing. Another species of vocal fish, the silver perch (Bairdiella chrysoura), is commonly called the drum or croaker for the repetitive throbbing or drumming sounds it makes. During the spawning season, male drums vocalize in groups, in a phenomenon known as fish chorusing. The result in both examples is a synchronized release of sperm and eggs, again illustrating the powerful biological role played by otolithic hearing in fish behavior.
By the time of amphibians during the Devonian era, about 380 million years before the present, inner ears had evolved additional non-otolithic hearing organs, most notably the basilar papilla and amphibian papilla. These function by a fluid-dynamical mechanism similar to the cochlea and thus give access to a wider range of acoustic frequencies than can be detected by otolith organs alone. With the additional apparatus of a tympanum and middle ear for hearing in air, these new sensors would provide landdwelling amphibians with the capacity to detect sounds from conspecifics and other sources at a distance in the terrestrial environment.
Nevertheless, the amphibian otolith organs, and especially the saccule, retain a remarkable sensitivity to seismic vibrations. This sensitivity is especially useful, of course, for species that return to water to reproduce during the spawning season. In the United States a familiar example of such a species is the American bullfrog (Rana catesbeiana). The frog returns every spring to a favored breeding pond for a few weeks for the purpose of spawning. As with fish, the male frogs do most of the vocalizing, attempting to attract females. The frog, of course, has both a larynx and a vocal sac, which it inflates to help radiate its sound, and the seismic sensitivity of its saccule allows it easily to detect surface vibrations produced by the vocal sac of other frogs.
Further up the vertebrate phylogenie tree are the archosaurians, comprised of crocodilians, dinosaurs, and birds. This group, which emerged from reptiles during the Permian era about 250 million years ago, shows many similarities, including nesting and parenting behaviors. In terms of the evolution of hearing, these animals share several interesting features: They all are highly vocal, tend toward chorusing behavior, and have anatomically similar inner ears. Archosaurians possess a structure that is called a basilar papilla, like the organ of the same name in amphibians, but in this subclass it functions more like the mammalian cochlea.
A striking example of vocalization within this group comes from the male American alligator (Alligator mississippiensis), whose bellow begins at such a low frequency (about 20 hertz) that it falls below the range of human hearing. Although we can’t hear 20-hertz component of the bellow, we can actually see the effect of it, because it is so powerful that it cause the water to dance on the alligator’s back. Meanwhile, the intended recipients of the male’s message can hear it quite well; the alligator saccule is anatomically large, and like the frog saccule it is well suited to detect the 20-hertz infrasonic component transmitted in water, as I demonstrated in a 2007 paper in Journal of the Acoustical Society of America.
Why Do Human Ears Have Otoliths?
When the mammalian cochlea is viewed in a comparative-anatomy framework alongside the basilar papilla in amphibians, reptiles, and archosaurians, one striking insight that emerges is that all these structures are in close anatomical proximity to the saccule. This proximity is consistent with the hypothesis held by a number of workers in the field that the mammalian cochlea evolved from an invagination within the saccule. The cochlea itself is an extraordinary organ: By means of mechanical properties of the basilar membrane and neurophysiological tuning of the hair-cells that are arrayed along it, with high-frequency sounds represented at its base and low-frequency sounds at the apex, the cochlea carries out a spectral analysis of sound. This capability, in turn, enables us to develop highly nuanced acoustical communication in the form of speech, to appreciate music, and to form complex mental representations of the environment-a capacity referred to as auditory scene analysis. But could the otolith organs in human have conserved an acoustical sensitivity as well?
The answer to this question is most definitely yes. Physiological experiments recording from the vestibular nerve in animals have shown that hair-cells within otolith organs respond to sound and vibration in much the same way as cochlear hair-cells. Otolith acoustic sensitivity in humans can be demonstrated noninvasively by means of evoked potentials, or electrical responses to sound or vibration. The measured electrical activity is usually obtained from surface electrodes placed over certain muscles around the eyes or on the neck that are associated with the vestibular reflex pathways, which help us keep our balance or maintain eye gaze when we are moving. These are referred to as VEMPs, vestibular evoked myogenic potentials. Experiments using airconducted sound stimuli reveal that the VEMP threshold for human otolith organs is quite high, but not that high: about 70 to 80 decibels of sound pressure level, roughly equivalent to the loudness of normal speech. Another property of VEMPs is that they are timed-that is, the responses are greatest around a certain frequency, in this case around 500 hertz (also in the range of the most sensitive frequencies in fish, as the reader will recall). If skull vibration rather than air-conducted sound is used as the stimulus, an additional low-frequency sensitivity of about 100 hertz becomes apparent. Under certain circumstances, a person undergoing a VEMP test may have such a great sensitivity to low-frequency skull vibration that he or she shows physiological responses even below the threshold of human hearing. Here, too, we see a similarity to the remarkable seismic sensitivity of the amphibian saccule.
The ease with which it is possible to record VEMPs has allowed audiologists and audio-vestibular medical professionals worldwide to use this technology as a tool for the diagnosis of vestibular dysfunction. With tire appropriate combination of stimuli and reflex pathways it is possible now to test the integrity of the human utricle and saccule independently. This capability, in combination with other newly developed methods, may soon make it possible for the first time to test each of the five human vestibular end-organs on its own: not only the otolith organs but also the three semicircular canals, which by virtue of their positioning at right angles to one another are highly sensitive to angular accelerations (rotation of the head). Remarkably, then, advances now taking place in vestibular medicine can be traced back to the rediscovery of an otolithic acoustic sensitivity which has been conserved from the hearing mechanism of our swampy ancestors.
An Accessory Auditory System?
Having established that the human inner ear has indeed conserved a primitive acoustic sensing mechanism, we can proceed to the question of why. What function does this mechanism serve in humans? Perhaps it has none; after all, whatever the role of the otolith organs in fish and frogs, it is entirely possible that with the development of the cochlea any remaining sound sensitivity in mammals could be simply vestigial. As discussed above, however, the sense of hearing carries out several important roles. One role is scene analysis, for which purpose the cochlea is an organ par excellence, refined by millions of years of natural selection. In contrast, it seems unlikely that the otolith organs, with their much simpler physiological properties, could add significantly to this function over and above that of the cochlea.
Let’s recall, however, that another function of hearing is to mediate courtship. Given that otolithic hearing mechanisms mediate this function in fishes and amphibians, is it possible they continue to fill this role in higher vertebrates, including humans?
Examples of sexual behavior, cited earlier, in fish and in frogs show clearly that courtship vocalizations have a profound biological impact on the receiving animal, enough to cause the release of eggs and sperm. This must mean that stimulating sounds cause direct changes in the receiving animal’s autonomic and endocrine systems. Biologist Timothy Neary and his colleagues, who have closely studied the pathways that may mediate these auditory-endocrine effects in frogs, have identified a mid-brain sensory relay that sends auditory information to the hypothalamus. In mammals, the counterpart of this pathway is likely to be a vestibular connection to another mid-brain nucleus, the parabrachial nucleus, which is shared with the visceral sensory-motor system. The signaling pathways of the parabrachial nucleus, in turn, reach not only to the hypothalamus (exactly as they should do if this nucleus is to mediate vocal-endocrine effects), but also to the amygdala, associated with emotion, the mesolimbic dopamine system, which is associated with reward, and higher affective centers, including the cingulate limbic system.
In fact, the vestibular-visceral-parabrachial system has been studied in humans for the role it plays in various negative symptoms of vestibular disorders, including motion sickness, vertigo and anxiety. Yet the same system appears to be responsible-in some people, at least-for the thrill and pleasure associated with vestibular activation from inertial or gravitational stimulation, for example in riding a roller coaster. Acoustic activation of the same system from sound or vibration, therefore, may well fit the bill as a mechanism for mediating an otolithic hearing function. Given the similarity in some of the central projections of the accessory olfactory system, an acoustic vestibular-visceral-parabrachial pathway can reasonably be thought to play a similar role-that is, as an accessory auditory system.
The case for an accessory auditory system in humans gains some support from the observation that vocal courtship behavior associated with movement-that is, dance-is ubiquitous throughout the vertebrate classes. The coupling of ritualized movement and sound is especially noticeable in birds, which, in addition to being highly vocal like their archosaurian cousins, also engage in mating dances, invariably accompanied by characteristic calls. The elaborate dances of the Japanese crane (Grus japonensis), the Western grebe (Aechmophorus occidentalis), and die Wandering albatross (Diomeda exulans) are just a few examples among many. Mammals, including primates, offer their own examples, such as the pair-bonded Siamang gibbons (Hylobates syndactylus), which perform elaborate vocal duets accompanied by acrobatic locomotor displays. This intimate sound/movement link associated with such courtship or pair-bonding behavior cannot be explained easily by a cochlea-based hearing system, but it can very naturally be explained by an otolithic accessory auditory system.
These examples offer good circumstantial evidence from among various vertebrates, but because our original question was framed in human terms, the strongest answer will come from evidence of human behavior. As is commonly noted, much human music is made to accompany dancing. Moreover, dance music tends to be loud, well above the vestibular threshold, and to contain a strong bass line at frequencies exactly within the region of maximum sensitivity of the vestibular system. Music with a solid rhythm or groove almost compels us to move in time with the beat. Recent evidence discussed by me and others in Frontiers in Human Neuroscience strongly suggests a considerable overlap between the areas in the brain activated by rhythm and parts of the vestibular brain network-in fact, one might even think of rhythm perception as a form of vestibular perception. It seems quite reasonable to speculate, therefore, that an accessory auditory system-one concerned with balance and rhythm rather than with complex acoustical analysis-may play a role in courtship and sexual behavior as a fundamentally important aspect of human life. Such an accessory auditory system may well have been central in the evolution of human music and dance.
Finally, if this all sounds a bit farfetched, it is worth noting that my hypothesis is entirely consistent with Charles Darwin’s own theory of the origin of music: that it evolved from vocal courtship behavior. On reviewing vocal behavior among a variety of species, from frogs, alligators, and birds to mice and gibbons, he concludes: “Unless females were able to appreciate such sounds and were excited or charmed by them, the persevering efforts of the males,.., would be useless; and this is impossible to believe.” After reviewing human music of different cultures, he reaches similar conclusions: “All these facts with respect to music and impassioned speech become intelligible …, if we may assume that musical tones and rhythm were used by our half-human ancestors, during the season of courtship….” Appearances notwithstanding, gyrating people at a dance party and shoaling fish at spawning season may have more in common than they realize.