Possible preadaptations to speech. A preliminary comparative

HUMAN EVOLUTION Vol, 19 - n. 1 (53-70) - 2004

Verhaegen M. Studiecentrum Antropologie, Mechelbaan 338, 2580 Putte, Belgium marc.verhaegen @village.uunet.be

Munro S. School of Archaeology and Anthropology Australian National University u3 3 72190 @ ana.edu.au

Keywords: Speech origins, language evolution, hominid diet, human evolution, aquatic theory, musical abilities, diving abilities, suction feeding, consonants, Homo erectus, seafoods, comparative biology.

Possible Preadaptations to Speech. A Preliminary Comparative

Human language is a unique phenomenon and its evolutionary origins are uncertain. In this paper we attempt to explo-

re some of the preadaptations that might have contributed to the origin of human speech. The comparative approach we use is based on the assumption

that all features of a species are functional, and that all features can be compared with those of other animals and correlated with certain lifestyles. Using this method we attempt to reconstruct the different evolutionary pathways of humans and chimpanzees after they split from a common ancestor. Previous results from comparative studies suggest human ancestors may not have evolved on the open African savan-

nas as was once believed, but more probably were coastal omnivores feeding on plant matter and easy to catch inverte-

brates such as shellfish from beaches and shallow waters. Fossil and archaeological data suggest this coastal phase occurred at the beginning of the Pleistocene, when Homo ergaster-erectus dispersed between East-Africa, North- Africa, South-Asia and Indonesia. This paper presents comparative data suggesting the various human speech skills may have had their origins at different times and may originally have had different functions. Possible preadaptations to speech include, for instance, musical skills present in a variety of primate species (sound production); airway closure and breath-hold diving for collec- ting seafood (voluntary breath control); and suction feeding adaptations for the consumption of fruit juice or certain seafoods (fine control of oropharyngeal movements). The different evolutionary pathways of chimpanzees and humans might explain why chimpanzees lack language skills and why human language is a relatively recent phenomenon.

54 VERHAEGEN, MUNRO

Introduction - Comparative Anthropology

Three major components of human language - phonology, semantics, and syntax -

are acquired successively from about the first, the second and the third year of life

(Hirsch-Pasek & Golinkoff 1996). This succession may reflect the human linguistic

evolutionary stages: pre-language, one-word sentences, and grammatical or 'true' lan-

guage. This paper discusses the first stage, the phonetic pre-adaptations for language, and is based mostly upon comparative data with other mammals.

In constructing human evolutionary models, paleoanthropologists tend to focus on

the fossil evidence, but the comparative method (comparing the anatomy, physiology, behaviour, DNA etc. of living animals) is probably more secure, systematic and reliable.

By using the comparative data we adopt an analytical andfunctionalistic approach.

Biological features are generally inherited independently of each other (Mendel's

Laws), due to the crossing-over and independent assortment of chromosomes during

gametic reproduction. This recombination of genetic material may not only explain how

features can evolve in parallel, i.e., apart from each other, but also why selection, wor-

king on different features in parallel in all members of a population, can be so efficient.

Since most biological features are polygenetic (influenced by more than one gene) they

can be 'fine-tuned' through the processes of recombination and selection. All species

have gone through an immense period of recombination and selection - not one of all

our millions of ancestors died before it had produced fertile offspring - therefore every

feature must have had one or more functions. These functions are not always obvious,

for instance, features can have multiple functions and functions can change over time

(evolutionary opportunism), but by comparing the similar features of different species

it is often possible to identify general trends and correlations between certain features

and particular lifestyles or environments.

The comparative approach can be used for the anatomical, physiological and beha-

vioural features of all animals. According to biomolecular data (DNA and proteins),

chimpanzees are our closest relatives. By comparing the features of humans and chim-

panzees, therefore, it may be possible to work back towards the Last Common Ancestor

of Homo and Pan, and attempt to reconstruct its likely habitat and behaviour. It may

also be possible to determine which features humans have since acquired, and the type

of environmental factors that may have been responsible for these acquisitions.

Perhaps the main drawback of the comparative approach is the insufficient data

available for the great majority of species. Many anthropologists, for example, are not

particularly familiar with the anatomies, physiologies and behaviours of non-primate

species, and have shown a preference for the hominid fossil evidence rather than the

POSSIBLE PREADAIrFATIONS TO SPEECH 55

comparative evidence. However, while the fossil record can provide additional insights,

its importance can also be overstated. Fossils are incomplete - typically they are frag- mented pieces of bone without soft parts - and are usually of uncertain relation to living

species. Frequently, species, age and sex are unknown, and sometimes the geological age and palaeo-environment are uncertain.

In this paper, we first briefly outline the results of our comparative studies and present a hypothesis for human evolution: a waterside scenario (Verhaegen & Puech 2000; Verhaegen et al. 2002). Then, more specifically, we compare the anatomy and physiology of the food and airways of humans, chimpanzees and other mammals, provide a

list of possible functions, and attempt to determine how and why the human peculiari- ties associated with speech evolved.

A Waterside Scenario

Most researchers agree that our remote primate ancestors lived in trees, but there is some disagreement as to how humans became independent of trees. Recently, there has been a steady accumulation of evidence suggesting that humans may not have evolved in a warm and dry environment as was once commonly believed, but instead may have evolved in warm and wet conditions, at the edge between land and water (Hardy

1960; Ellis 1991; Morgan 1997; Bender et al. 1997; Tobias 1998; Verhaegen et al.

2002). Anatomical, physiological, biochemical and palaeo-environmental data tend to support this view. At the 1999 symposium on Water and Human Evolution, in Ghent, Belgium (Vaneechoutte 2000), it was proposed that our ancestors were coastal or river- side omnivores who not only consumed terrestrial plants and animals, but also collec- ted food from shallow waters. This scenario is supported by a comparison of some typically human features with those of other mammals.

Bipedality: climbing-wading origin? Primates, perhaps because they are tradi- tionally climbing animals, have a tendency to adopt an erect posture, and this behaviour is accentuated in primate species that frequently wade through shallow water. Proboscis

monkeys, for example, cross shallow stretches of water on two legs when moving from

one mangrove tree to another (Napier & Napier 1967). Lowland gorillas wade on their hindlimbs through forest swamps in search of wetland plants and sedges (Chadwik 1995; Doran & McNeilage 1997). Dental microwear and isotopic evidence suggest the australopithecine diet may have also included such plants (Puech et al. 1986; Puech 1992; Sponheimer & Lee-Thorp 1999). Therefore it is likely that Pliocene hominids also regularly waded bipedally in the shallow waters of forest clearings, gallery forests

56 VERHAZCEN, MUNRO

or mangrove areas, possibly in search of floating fruit, wetland plants, reed sedges, fish

or shellfish (DuBrul 1977; Ellis 1991; Broadhurst et aL 1998; Verhaegen 1992;

Verhaegen et al. 2002). This is corroborated by the recent discoveries of the 7- to 6-million year-old, probably bipedal-and-climbing, hominids Sahelanthropus tchadensis and

Orrorin tugenensis in shallow perilacustrine environments (Brunet et al. 2002; Senut et

aL 2001). - Thick enamel and stone tool use: hard-shel led f oods? A combination of thick

molar enamel and stone tool use is known to have existed in various H o m o species, and exists today in capuchin monkeys and sea otters. Sea otters have large, flat cheekteeth,

which resemble those of australopithecines (Walker 1981), and use stones to crack open

shellfish while floating on their backs. Capuchins open nuts with stones and use oyster

shells to remove shellfish from the tnmks of mangrove trees (Fernandes 1991).

Chimpanzees, which have thinner molar enamel, manipulate stones to crack open hard- shelled nuts. Human Pliocene ancestors, perhaps in the same way as mangrove capu-

chins, might have used stones or other hard objects to remove coconuts from palm trees,

to crack hard nutshells, or to remove and open oysters from the trunks of mangrove

trees (Verhaegen et al. 2002). During the late Pliocene or early Pleistocene, members of the genus H o m o - as

opposed to our more distant relatives the australopithecines - might have also learned

how to duck the head underwater and to dive and collect underwater shellfish as well

as other aquatic resources. Humans have much more efficient diving capabilities than

nonhuman primates (Schagatay 1996; Morgan 1997; Bender 1999; Verhaegen et al.

2002). Indeed, Homo fossils - as opposed to australopithecines - are typically found

near shellfish (Chiwondo, Chemeron, Nariokotome, Zhoukoudian, Boxgrove, Terra

Amata, Rabat, Hopefield, Gibraltar and others). Although sea level rises and the actions

of tides and waves have drastically reduced the chances of discovering hominid fossils

at sea beaches, H o m o erectus remains have been discovered amid shellfish, barnacles

and corals, from the early Pleistocene skull of Mojokerto at Java (Ninkovich & Burckle

1978), to the late Pleistocene Acheulean tools of Eritrea (Walter 2000; Walter et al.

2000). Stone tools discovered on Flores suggest H o m o erectus crossed a 19 km wide,

deep oceanic channel more than 800,000 years ago (Morwood et al. 1998; Tobias

1998). We have argued that the fast dispersal of H o m o erectus at the beginning of the

Pleistocene between Algeria (Aln-Hanech), and Java (Mojokerto) occurred along the

Mediterranean and Indian Ocean coasts, where foods could be gathered from both the

land and sea (Verhaegen et al. 2002). From the coasts, different H o m o sidebranches

could have migrated up rivers into the interiors of Africa and Eurasia, where fossilisa-

tion chances may have been more likely. Initially restricted to the edges of rivers,

POSSIBLE P~ADAPTATIONS TO SPEECH 57

swamps and lakes, some Homo populations later moved to areas further from perma-

nent water. Whereas stone tool use for cracking hard-shelled foods may have been a preadap-

tation for the development of lithic technologies, the diving abilities of our ancestors might have been a preadaptation for the development of voluntary speech (Morgan

1997; Diller 2000). Like Darwin (1871), however, we believe human sound production probably has

deeper roots, beginning at a time when our ancestors were still arboreal. There may be several overlapping preadaptations for speech, including musical abilities, swallowing

abilities, the ability to close the airways, the ability to control breathing, and the ability to communicate symbolically.

A Short Survey of Food- and Airway Adaptations in Mammals

In most animals the mouth is used primarily for feeding, and most oropharyngeal

adaptations are directly linked to feeding behaviour (foodway). In amniotes (reptiles, birds, mammals) the nose is primarily for breathing (airway). Sound production outsi- de the water is normally linked to the airway, but for the production of loud sounds a wider space may be advantageous. This is possibly why singing birds and barking dogs make a connection between the airway and the mouth. The traditional functions of the mouth, nose and throat cavities are discussed here under three headings: air, sound and food.

1. AIR On land: All land mammals breathe air through their nostrils, though some can also

inhale and exhale through their mouths. A number of hypotheses have been put forward

to explain the evolutionary function of the human external nose. For example, it has been argued that the human nose is designed to prepare air for breathing: to purify, moi- sturise, filter or warm it, or to retain water from expired air, etc. (Franciscus & Trinkaus

1988a, 1988b). There are, however, no comparative examples of mammals developing an external nose for similar reasons. The only other primate with a well developed external nose is the mangrove-dwelling proboscis monkey, which is well known for its

swimming ability and can swim several metres under water (Napier & Napier 1967). Most land mammals, as opposed to humans (except babies), have their larynx posi-

tioned high in the throat. At rest, the larynx connects with the nasal passage, its entran- ce well within the nasal cavity, and acts as a barrier separating the nasal passage from

5 8 VERHAEGEN, MUNRO

the oral cavity. This means most mammals can swallow fluids (and in some species semi-solid materials) and breathe simultaneously (Laitman 1985; Crompton et aL

1997). Some mammals, like red deer and koalas, have a permanently low larynx, but

have evolved (at least in red deer) a long velum which connects the nasal cavity with the larynx when at rest. Humans have a descended larynx, like koalas and red deer, but lack an elongated velum. Thus, while most mammals, including human babies, can swallow fluids and breathe simultaneously, humans above the age of about six months cannot.

In water: Diving mammals must be able to close their airways underwater. They

must also have adaptations that allow for the considerable water pressure that can be

placed upon air-filled cavities (middle ear, sinuses, bronchi, lungs) and for the great and sudden pressure changes that can be experienced while diving and surfacing. Moreover, they must be able to inhale large amounts of air rapidly when they surface in order to minimise the length of time between dives. It has been argued that some semi- or ex- aquatic mammals, such as tapirs, elephants, hooded and elephant seals, initially evolved

elongated external noses to help them breathe and to prevent water from entering the airways while wading or swimming (the human nose seems well designed hydrodyna- mically to keep water out of the airways while swimming at the surface or under water or while dipping the head under water or diving into water). Most aquatic mammals normally breathe through the nostrils (the whales' blowhole), although some, like walruses, frequently mouth-breathe (Fay 1982).

Olfaction: Most mammals use the air they breathe for olfaction. In land mammals the sense of smell is very important, but in many aquatic mammals this function has been reduced and in some cases completely lost (Dehnhardt 2002).

Panting: The rapid forcing in and out of air through the mouth is a method used by

many mammals (including humans when thermoregulatory sweating does not work well, for instance, in very humid conditions) to help reduce body temperature. In dogs, the velum (soft palate) opens during inhalation and closes during exhalation with each pant cycle (Schmidt-Nielsen & Taylor 1970; Biewener et aL 1985).

2. SOUND Calls: Sound production is derived from the function of breathing. Instinctive and

territorial sound productions, like barks and roars, are seen in many mammals; humans cry and laugh aloud when stimulated. Many of these 'automatic' sounds are reflexes

controlled by neural centres in the brain stem. Most are produced by the glottis slit between the vocal folds in the larynx. The lower the larynx during sound production, the

POSSIBLE PREADAPTATIONS TO SPEECH 59

louder and more impressive the territorial calls (Fitch 2000; Ohala 2000). An extreme example is the male hammerhead bat, which has an extremely large and low, in fact intrathoracal, larynx (Rosevear 1965). Monkeys, dogs, pigs and all other mammals exa-

mined so far, lower the larynx and close the velum during loud vocalisations. In male red and fallow deer, during roaring, and probably also in other mammals during loud

calls, the vocal tract has a horizontal (oral) and vertical (pharyngeal) tube, which is reminiscent of the permanent situation in humans. This two-tube configuration is said to have implications for sound production, the expanded pharynx perhaps allowing

humans to produce the full range of speech sounds (Laitman 1985). Many primates, including all hominoids except humans and smaller gibbons, have large laryngeal airsacs, possibly for making their long calls louder or faster or for preventing hyperventi- lation (Hewitt et al. 2002; Ankel-Simons 2000).

Song: Musicality in animals is often correlated with an arboreal (many birds, some

primates) or aquatic (some cetaceans and pinnipeds) lifestyle. The songs of humpback whales are particularly well-known. Primate examples include the pant-hoots of chimpanzees. Male proboscis monkeys, who have longer external noses than females and infants, are said to use the nose as an organ of resonance in vocalization (Ankel-Simons 2000), and even to produce a typical double sound through the nose and mouth at the

same time (Napier & Napier 1985). More elaborate songs are seen in some monogamous primates such as indris, tarsiers, titi monkeys and gibbons (Darwin 1871; Vaneechoutte & Skoyles 1999; Miiller & Anzenberger 2002). At least in primates, these sounds appear to be largely under the emotional control of the limbic cortex (Deacon 1997). The songs of birds, as well as mammals, differ according to the population (dialects). Birds typically learn their songs from their fathers during a sensitive period early

in life (presong). Speech: Speech production is uniquely human, though some birds, such as moc-

kingbirds and parrots, can mimic human speech (Pepperberg 2000). Human speech

sounds are produced by muscles under the voluntary control of the greatly enlargedpre- central cortex (Brodmann's Area 4, the motor cortex that controls fine skeletal muscle

movements). Consonants are produced by using the lips, tongue, jaw, velum, pharynx and glottis, although in birds (with inflexible beaks and reduced tongues) these sounds are probably imitated by muscle contractions in the syrinx, the vocal organ, where the right and the left bronchus come together. The only mammals known so far that can mimic human utterances, presumably without any understanding of the meaning, was a

harbour seal that had learned to produce (albeit with a throaty voice) fragments of humanlike speech from a fisherman, probably at a sensitive period early in its life (Rails

et al. 1985; Deacon 1997), and to a very limited extent a beluga whale (Eaton 1979).

60 VERHAEGEN, MUNRO

Very variable sounds such as clicks, though not humanlike, are produced by dolphins in the nasal air passages, probably mainly by using the larynx, which, unlike the human larynx, is very 'ascended', i.e. permanently locked in the nasal cavity (Slijper 1958; Dudok van Heel 1970).

3. FOOD Biting: The anatomy of the oral cavity in mammals is probably influenced mainly

by its primary function, the processing of food through biting, chewing and swallowing. Many fruit-eaters have spatulated incisive teeth, as do most primates including humans, whereas these are more conical in meat- or insect-eaters (for primates see Ankel-Simons 2000), and can be reduced in terrestrial and aquatic plant-eaters. Terrestrial carni-, insecti- and omnivorous mammals typically have long canines, whereas pure herbivores, including the aquatic sirenians, have reduced or absent canine teeth. Aquatic carnivores such as cetaceans, and to a lesser degree pinnipeds, generally do not have very long canines. All their teeth (front teeth and cheekteeth) tend to be sharp and conical, probably for catching slippery fish or squid. An obvious exception is the walrus, which has long canine tusks probably for intraspecific display (e.g. de Muizon 1995). Large front teeth can be important for attacking sexual rivals or as a defence against predators, or as a warning signal. It has been argued that showing the front teeth, as is seen in human laughter, can indicate the physical strength, health and self-confidence of an individual.

Chewing: Premolars and molars are for processing the food, for instance, cracking hard foods with thick enamel, chewing calorie-poor plants with fiat cheekteeth, cutting through tough plant food or insect exoskeletons with sharp ridges of thin enamel, or sli- cing meat with the carnassial teeth. Typically, mammals such as ungulates and carnivores, as well as most primates except humans, have low, long and horizontal palates with transversal ridges of comified epithelium, probably for fixing food while chewing (Romer & Parsons 1977).

Swallowing: As already discussed, in most mammals the larynx at rest is engaged in the nasal cavity, which means that swallowing fluids (mouth to esophagus) and breathing (nose to larynx) at the same time is possible. A permanently lowered larynx makes this impossible for humans, except babies. Not all kinds of food have to be chewed tho- roughly. As opposed to herbivores, carnivores often swallow large food boluses, which seems to require a large pharyngeal space (laryngeal descent creates a larger pharyngeal space, see figure 1). Foods such as fruit juice and tree exudate (some bats and primates) or insects and grubs (ant-eaters, bears) can be sucked and swallowed without chewing.


Some juice- or sap-sucking New World monkeys show features that are believed to increase the suction drainage force, such as an angular and highly vaulted palate in

some marmosets and a humanlike closed upper dental arch with incisiform canines in dusky titis (Hill 1957; Hershkovitz 1977; Jones & Anderson 1978). Many pinnipeds, most notably walruses and bearded seals, have round tongues and smooth vaulted palates, probably to enable them to more easily suck the smooth and slippery seafoods out of shells (King 1972; de Muizon 1993). Most land mammals, including nonhuman pri-

mates, and most aquatic mammals, including hippos, otters and furseals, have palatal ridges, but these are lacking in some cetaceans, crabeater seals, elephant seals and wal-

ruses (Roger Crinion, personal communication). The absence of palatal ridges may allow food to slide through the mouth unchewed. The strong suction feeding of walruses probably requires a round and relatively narrow oral cavity and a strong retractable

tongue, and possibly an enlarged pharyngeal space and lowered larynx (Fay 1982; de Muizon 1993). Most aquatic mammals can swallow food underwater, although it is not clear which specific adaptations make this possible. In many aquatic mammals such as walruses, sealions and seacows, the epiglottis, the lid that covers the well-developed larynx of humans during swallowing and prevents food entering the trachea, is not as well developed as it is in humans, monkeys, pigs and probably most terrestrial mammals (Negus 1949).

Fig.1 Midsagittal Section through the Heads of a Chimpanzee and a Human (after Laitman 1985). Note the protruding nose and chin, the round tongue and short, vaulted, smooth palate, and the lowered larynx in humans.

6 2 VERHAEGEN, MUNRO

Fig. 2. Possible Function of Upper Lip and Philtrum: Closing the Nostrils (after Morgan 1997). Note

that human ancestors were more prognathous, and possibly had moustaches, so that the upper lip was

closer to the nostrils.

Human feature External nose Inferior nostrils Smaller mouth Vauhed palate Round tongue Smooth palate Very low larynx Mobile larynx Flexible glottis Mouth breathing Baby babbling

+ additional description long, narrow nose passage human philtrum in upper lip with red everted lips with parabolic tooth row fits nicely in palate less transversal ridges no intranarial engagement large pharyngeal space very muscular vocal folds under volitional control song: sensitive period

Possible function(s) originally Keeping water out of airways? Sexual selection?? Resonance?? Keeping water out? Nose closure (figure 2)? Acoustic?? Mouth closure. Suction of fruits? of seafood? Sexual selection?? Food suction, e.g. fruits? seafood? Underwater?? Chewing & swallowing. Food suction? Vocalisation? Swallowing smooth and slippery seafood? Suction? Vocalisation? Suction? Large intake of breath? Choking danger! Vocalisation? Suction? Swallowing large food boluses? Singing (tone height)�9 Glottis closure for diving?? Diving? Singing? Speaking? Singing? Speaking?

Table 1 -- Nose-Mouth-Throat Differences with Chimps

Abilities

Human vs. chimp features

Animal examples

Secondary use in speech

Song? �9 Tone & rhythm �9 Duetting �9 vocal folds very muscular �9 sensitive period & babbling ? lower larynx & louder calls ? mobile larynx ? no laryngeal airsacs �9 Musicality: many birds;

gibbon, tiff, tarsier, indri; humpback whale

�9 Speech imitation: some birds, harbour seal

�9 Tone, intonation �9 Rhythm �9 Dialogue �9 Sound imitation

Seafood? Smooth & slippery Swallowing underwater??

? small mouth & fleshy lips ? smooth palate with less ridges ? vaulted palate & round tongue ? mobile hyoid & larynx ? large pharynx �9 Suction feeding: sloth bear;

walrus & other pinnlpeds; some primates & bats

Extreme lip, tongue & pharynx control; dental, palatal etc. closures easy (not only labial)

Click sounds

Diving? Voluntary breathing Airway closure

? small mouth & everted lips ? ph/ltrum & slitlike nostrils ? long & narrow nasal passage .9 round tongue, smooth palate ?? laryngeal descent �9 External nose: proboscis

monkey, tapir, elephant, hooded & elephant seal

Airway closure & voluntary breathing in diving mammals

Breathing muscles: voluntary in- & expiration

Airway closures at lips, tooth row, palate, velum, glottis...

Table 2 -- Possible Explanations of the Human-Chimp Differences


Possible Explanations for the Human-Chimp Differences

Chimpanzees lack an external nose, slittike nostrils and aphiltrum (the vertical fur-

row in the human upper lip), and have a shorter and more direct air passage from the nostrils to the nasopharynx than humans (see figures 1 and 2). In humans the nasal air passage is both longer and narrower and has an inverted U shape (with the nostrils underneath the nose instead of in front).

Chimpanzees have a larger and more protruding mouth (prognathism) with larger

canine teeth and corresponding gaps (diastemas) in the opposite jaw, whereas humans have a smaller mouth with everted tips, a closed and parabolic tooth row and teeth of nearly equal height (Hocket 1967; Laitman 1985).

Chimps have a fiat tongue and a long and transversally ridged palate, whereas humans have a round, thick and bulbous tongue, a much-shortened oral space and a short, wide, deep palatal 'vault'. The human tongue can be shaped to fit tightly against the arched and smooth palate.

In chimpanzees, the gap between the palate and larynx is smaller than in humans, who have a tongue bone (hyoid) and larynx "retreated still farther down in the neck" (DuBrul 1958). Humans have a well-developed larynx and very muscular vocal folds,

but lack airsacs (Negus 1949). Humans have a very large representation of the oral muscles in Area 4 (see above),

and only in humans does damage of Area 4 produce muteness (Deacon 1997). Humans, unlike chimpanzees and other primates, have an Area 4 representation of the larynx and

breathing musculature, have direct fibers connecting Area 4 to the nucleus ambiguus (cortico-ambiguus connections), and can voluntarily control the larynx muscles (nucleus ambiguus) and the breathing muscles (brain stem).

1. SINGING - VOCAL FOLDS Babies of two or three months produce cooing sounds. This is called vocalising and

is performed with the vocal folds in the larynx, without much oral movement. Soon the- reafter, even in deaf children, the babbling starts to include labial consonants, and sylla-

bles are produced (consonant plus vowel). In babies older than six months, the sound pattern already resembles the native language, and 'dialogues' with the mother stimu- late the utterances. This sensitive period of automatic sound production and the learning of local dialects resembles the subsong period in birds (Deacon 1997).

The early prelingual sounds, without symbolic meaning, may correspond with the elaborate songs of nonhuman primates like gibbons. Music powerfully affects the emo-

64 VERHAECEN, MUNRO

tions (anthems, hymns, marches, love songs), and is used by humans as a territorial and

pair- or group-binding behaviour, as it is by other animals. Well-developed musical abi-

lities and duet-singing are seen in monogamous primates such as gibbons. Bonobos

(pygmy chimpanzees) engage in group-chorusing and rival males have been observed

engaging in vocal duels (De Waal 1997). Some aquatic species such as humpback wha-

les (polygynous) also use complex melodic utterances for territorial behaviour.

Interestingly, the harbour seal raised by a fisherman only 'spoke' in a humanlike way when it was not engaged in emotional or territorial behaviour (Deacon 1997). It is

known that musical training in young children induces an enlargement of the temporal

and insular cortex (planum temporale) in the left brain hemisphere, and can lead to a n

improvement in a child's ability to hear absolute tones (Schlaug et al. 1995). Intonation is an indispensable element of all spoken languages, and almost half the world's lan-

guages are tonal. Ohala (2000) argues that our descended larynx could not have evolved as an adap-

tation for speech, since men, who have an even lower larynx than women, are not bet-

ter adapted for speech than women, who perform better in verbal tests. In fact, the com-

parative evidence suggests that laryngeal descent may have occurred in order to make

the voice more impressive (Fitch 2000; Ohala 2000). This, however, does not explain

why the human larynx is incapable of making a direct connection from the nasal passa-

ge to the larynx and cannot engage in the nasopharynx (intranarial or suprapalatal), as

it does in human babies and most land mammals.

2. DIVING - AIRWAYS At birth, humans have a larynx that can engage in the nasopharynx, then, between

four and six months, the larynx begins to descend. One possible explanation for laryn-

geal descent in humans could be the need to breathe a large amount of air in a short

period of time to facilitate diving (Morgan 1997). Whales and dolphins have perma-

nently intranarial larynges, ascended rather than descended.

All humans, probably unlike most nonhuman primates, can easily learn to dive, and

several human populations, such as some Indonesian and Oceanic populations, as well

as the Ama of Korea and Japan, collect shellfish through breath-hold diving (Schagatay

1996). Diving, as seen in aquatic or semi-aquatic mammals, requires a voluntary control

of breathing. In contrast with land mammals, divers must be able to take a deep breath

just before they intend to dive, and breathe deeper and faster between dives. They must


also be able to hold their breath underwater, paradoxically at the very time when their oxygen needs are highest. In contrast, terrestrial mammals intensify their breathing when they need more oxygen - while running, for example. This may explain why

humans are unique among primates in being able to control the larynx and breathing musculature at will.

Diving also requires the complete closure of the airways underwater, so water can be kept out of the lungs. Humans can close the oral (e.g. small mouth with fleshy tips,

closed tooth row, bulbous tongue and smooth palate) and nasal passages (e.g. slittike nostrils, longer and narrower airway) much more completely than chimpanzees can (figure 1). It has been suggested that human ancestors might have been able to close the nostrils by pressing the upper tip (moustached or not) against the nostrils as some peo-

ple do today when they dive (figure 2). The upper lip with the human philtrum seems "perfectly made for this. The two lines descending from the nose plug up the holes, while the recess in the middle allows for the bridge between the holes" (Peter van de Graaf, in Morgan 1997).

The inverted U shaped nasal passage in humans may have helped to keep water out of the airways while ducking the head underwater.

3. FEEDING - MOUTH AND TONGUE It has also been suggested, on the basis of comparative evidence, that our perma-

nently descended larynx, incapable of engaging in the nasopharynx, might have been a useful adaptation for suction of certain slippery foods or juices (Roger Crinion, personal communication). Animals featuring a descended larynx include not only some deer

and koalas that produce loud calls, but also so-called suction feeders such as some bats and perhaps some sap-feeders. It is possible that laryngeal descent allows considerable retraction of the hyoid and tongue so that the pressure in the oral cavity can be lowered, which is one possible way to accommodate underwater suction, as in walruses, or for

sucking juicy fruits, as is seen in some bats (Sprague 1943; Rosevear 1965; Hildebrand

1974; Fay 1982). Other adaptations seen in mammals that regularly suction-feed are a small mouth,

a smooth and vaulted palate, and a smooth and round tongue that can be shaped to fit tightly against the palate, as well as a closed parabolic upper tooth row without long canines and diastemas (gaps in the tooth row would hinder suction). These features, in different combinations, are seen in sloth bears, some bats and primates that suck insects,

fruit pulp or exudates, and in particular in walruses and other pinnipeds that include

66 VERHAEGEN, MUNRO

shellfish, squid or fish in their diet. These features also typically distinguish humans from apes.

The human sucking adaptations could have been used for fruits and/or for smooth aquatic foods. Humans do not have to chew raw oysters in order to eat them, and are

able to swallow small fish whole. Moreover, humans can swallow food underwater, and can also keep their mouths open underwater without swallowing or inhaling water. Feeding underwater requires a fine co-ordination of the lips, mouth, tongue and throat in order to keep water out of the airways and, at least in marine environments, to prevent ingestion of too much seawater. The human tongue is extremely flexible and is

well adapted for manipulating objects within the mouth. It is also well designed to help expel water from the mouth.

Some of the mouth-closing and/or feeding adaptations might explain why the human tongue is able to close the oral cavity at different places, allowing a diverse number of consonants to be produced, for example, at the alveolar, palatal, velar and uvular articulation places. The only nonhuman mammals, as far as we know, that are able to

reproduce recognisable pieces of human speech are the harbour seal (Rails et al. 1985) and possibly a beluga whale (Eaton 1979).

Concluding Remarks

The combination of comparative and fossil data suggests that by about 1.8 million

years ago human ancestors may have become more reliant on wading and diving than on climbing.

A waterside mammal might be expected to have greater control of the lips, tongue and throat muscles for seafood consumption, as well as voluntary control of the airway

and breathing musculature for swimming and diving. This oral cavity and airway control might have been preadaptative to the evolution of human speech, particularly in combination with the already well-developed rhythmical, melodic and duetting abilities

of our primate ancestors. A wading-and-diving lifestyle might have also required a different method of com-

munication. Traditional primate communication systems such as smell and certain types

of body language (such as posture, for example, though not facial expression) may have been less effective in a semi-aquatic milieu when compared to a purely terrestrial or arboreal one (Morgan 1997). Derek Ellis (personal communication) notes "how well

sound travels over water, compared to being muffled in forests, and even compared to


grassland. Foraging beach and lagoon apes could separate quite widely and still remain

in contact by vocalising."

It is possible that the modifications to our ancestors' food and airway entrances

coincided with an early stage in the disproportionate expansion of the human neocor-

tex, in particular Area 4 (precentral) and Area 44 (Broca), which control the fine move-

ments of the mouth and throat muscles - whether for singing, swallowing or diving.

Humans, as opposed to chimpanzees and other primates, have disproportionally large

neocortical areas when compared to the brain stem (e,g. Deacon 1997). Of these, the temporal and insular areas (including the Areas 4, 44 and Wemicke), where sounds are

produced, processed and interpreted, seem to have undergone the greatest enlargement

(Semendeferi & Damasio 2000). Perhaps in this part of the brain, the pre-existing func-

tions of song production, food consumption and airway control were integrated into a

system that could produce voluntary and articulated sounds, i.e. the beginnings of speech. The integration of this voluntary sound production system with the symbolic

powers that may have already existed in primates (Savage-Rumbangh 1986), might

have been made possible due to the extra brain tissue (association or integration cor-

tex) that developed during human evolution.

ACKNOWLEDGEMENTS - We are greatly indebted to Tecumseh Fitch, Elaine Morgan and Roger Crinion for information, corrections and ideas.

References

Ankel-Simons, E 2000. Pr/mate Anatomy. Academic Press, San Diego. Bender, R. 1999. Die evolutionsbio~g~" che Grundlage des menschlichen Schwimmens, Tauchens und Watens.

Dissertation, University of Bern, Switzerland. Bender. R., Vedaaegen, M. & Oser, N. 1997. Der Erwerb menschlicher Bipedie aus tier Sicht der aquatic ape theory.

Anthropologischer Anzeiger, 55, 1-14. Biewener, A., Soghikian, G & Cromtxon, A. 1985. Regulation of respiratory airflow during panting and feeding in

the dog.Respiratory Physiology, 61, 185-195. Broadhurst, C. L., Ctmnane, S. C. & Crawford, M. A. 1998. Rift Valley fish provided brain-specific nutrition for

early Homo. British Journal of Nutrition, 79, 3-21. Bnmet, M., Guy, E, PUbeam, D., Mackaye, I-k T., Likius, A., Ahounta, D., Beauvilain, A., Blondel, C., Bocherens,

H., Boisserie, J.-R., De Bonis, L., Coppens, Y., Dejax, J., Denys, C., Duringer, P., Eisenmann, V., Fanone, G, Fronty, P., Geraads, D., Lehmann, T., Lihoreau, E, Louchart, A., Mahamat, A., Mercemn, G, Mouchelin, G, Otero, O., Pelaez Campomanes, P., Ponce De Leon, M., Rage, J.-C., Sapanet, M., Schuster, NL, Su&e, J., Tassy, P., Valenfin, X., Vignaud, P., Vlriot, L., Zazzo, A. & Zollikofer, C. 2002. Anew hominid from the Upper Miocene of Chad, Cenlral Africa~ Nature, 418, 145-151.

Chadwik, D. 1995. Ndoki - last place on ear~ National Geog~hie, 188, 2-43.

68 VERrI~CEN, MUNRO

Crompton, A. W., German, R. Z. & Thexton, A. J. 1997. Mechanisms of swallowing and airway protection in infant mammals (Sus domesticus and Macaca fasicularis). Journal of Zoology, 241, 89-102.

Darwin, C. 1871. The descent of man. Appleton, New York. Deacon, T. W. 1997. The symbolic species. Norton, New York. Dehnhardt, G 2002. Sensory systems. A. R. Hoelzel ed. Marine mammal biology. Blackwell Publishing, Oxford. de Muizon, C. 1995. Walrus-like feeding adaptation in a new cetacean from the Piiocene of Peru. Nature, 375, 745-

748. De Waal, E 1997. Bonobo: the forgotten ape. University of Califomia Press. Diller, K. 2000. Why linguists and anthropologists should be i n ~ in the aquatic ape theory for the origin of

speech. J-L. Dessalles & L. Ghadakpour eds. The evolution of language. Ecole Nationale Sup6dowe des T616communications, Paris.

Doran, D. M. & McNeilage, A. 1997. Gorilla ecology and behaviour. Evo/ut/onary Anthropology, 6, 120-130. Dudok van Heel, W. H. 1970. DoO~jn, hoe doe je het ? Van Kampen, Amsterdam. DuBrul, E. L. 1958. Evolution of the speech apparatus. Charles C. Thomas Publications. DuBrul, E. L. 1977. Early hominid feeding mechanisms. American Journal of PhysicalAnthropology, 47, 305-320. Eaton, R. L. 1979. A beluga whale imitates human speech. Cam&ore, 2, 22-23. Ellis, D. 1991. Is an aquatic ape viable in terms of marine ecology and primate behaviour? M. Roede, J. Wmd, J.

Patrick & V. Reynolds eds. The aquatic ape:fact orfiaion? Souvenir, London. Fay, E H. 1982. Ecology and biology of the Pacific walrus Oabbenus rosmarus d/vergens. North American Fauna,

74, 1-279. Femandes, M. E. B. 1991. Tool use and predation of oysters (Crassostrea rhizophorae) by the tutted capuchin, Cebus

apeUa apella, in brackish water mangrove swamp. Pr/mates, 32, 529-531. Fitch, T. 2000. Vocal production in nonhuman mammals: implications for the evolution of speech. J-L. Dessalles &

L. Ghadalqgour eds. The evolution of language. Ecole Nationale Sup&ieure des T616communications, Paris. Franciscus, R. G & Tfinkaus, E. 1988a. The Neanderthal nose. American Journal of PhysicalAnthropology, 75,

209-210. Franciscus, R. G & Tnnkaus, E. 1988b. Nasal morphology and the emergence of Homo erectus.American Journal

of Physical Anthropology, 75, 517-527. Hardy, A. C. 1960. Was man more aquatic in the past? New Sc/ent/st, 7, 642-645. Hershkovitz, P. 1977. New World monkeys (Platyrrhini), Volume 1. University Press, Chicago. Hewitt, G, MacLamon, A. & Jones K. E. 2(102. The functions of the laryngeal air sacs in primates: a new hypothe-

sis. Folia Primatologica, 73, 70-94. Hildebrand, M. 1974. Ana/ys/s of vertebrate structure. Wiley, New York. Hill, W. C. O. 1957. Primates: Comparative anatomy & taxonomy, Volume IlL" Hapalidae. University Press,

Ecrmburgh. Hirsch-Pasek, K. & Golinkoff, R. M. 1996. The origins ofgranemar. MIT, Cambridge, Mass. Hocket, C. E 1967. The foundations of language in man, the small-mouthed animal. Scientific American, 217, 141-

144. Jones, C. & Anderson, S. 1978. Callicebus moloch. Mammalian Species, 112, I-5. King, J. 1972. Observations onphocid skulls. R. J. Harrison ed.Anatomy ofmar/ne mammab, Volume/.Academic

Press, New York. Laitrnan, J. T. 1985. Evolution of the hominid upper respiratory a'act: the fossil evidence. P. V. Tobias ed. Hominid

evolution: pas~ present and future. Liss, New York.


Morgan, E. 1997. The aquatic ~ t h e s / s . Souvenir, London. 392 Morwood, M. J., O'Sullivan, P. B., ~ E & Raza, A. 1998. Fission-track ages of stone tools and fossils on the east

Indonesian island of Floles. Nature, 392, 173-176. Miiller, A. E. & Anzenbelger, G 2002. Duetting in the tiff monkey Callicebus cupreus: structure, pair spedfidty and

development of duets.Fo/~a Primatologica, 73, 104-115. Napier, J. R. & Napier, E H. 1967. A handbook oftving primtes. Academic Press, New York. Napier, J. R. & Napier, E H. 1985. The natural history of the prbnates. M1T, Cambridge, Mass. Negus, V. E. 1962. The comparative anatomy and physiology of the larynx. Hafner, New York. Ninkovich, D & Burckle, LH 1978. Absolute age of the base of the hominid-beating bed in Eastern Java. Nature,

275, 306-308. Ohala, J. J. 2000. The irrelevance of the lowered Larynx in modern Man for the development of speecl~ J-L. Dessalles

& L. Ghadakpo~ eds. The evolution of language. Ecole Nationale Sup&ieure des T61&,,ommtmications, Paris. Puech, P-E 1992. Microwear studies of early African hominid tee~ Scanning Microscopy, 6, 1083-1088. Puech, P-E, Cianfarani E & Alberlini H. 1986. Dental microwear features as an indicator for plant food in early

hominids: a preliminary study of enamel. Human Evolution, 1,507-515. Pepperberg, L M. 2000. The Alex studies: cognitive and communicative abilities of grey parrots. Harvard University

Press, Cambridge, Mass. RaUs, K., Forelli, P. & Gish, S. 1985. Vocalizations and vocal mimicry in captive harIx)r seals, Phoca vitulina.

Ctmadian Journal of Zoology, 63, 1050-1056. Rorner, A. S. & Parsons, T. S. 1977. The vertebrate body. Saunders, Philadelphia. Rosevear, D. R. 1965. The bats of West Africa. British Museum, London. Savage-Rumbaugh, S. 1986.Ape languages:fromconditionedresponseto symbol. Columbia University Press, New

York. Schagatay, E. 1996. The human diving response: effects of temperature and training. Dissertation, University of

Iamd, Sweden. Schlaug, G, J'~incke, L., Huang, Y. & Steinmetz, J. 1995. In vivo evidence of saucatral brain asymn~Iry in musi-

clans. Science, 267, 699-701 Schrnidt-Nielsen, K. B. W. & Taylor, C. R. 1970. Panling in dogs: unidirectional airflow over evaporative surfaces.

Science, 169, 1102-4. Semendeferi, IC & Damasio, I-L 2000. The brain and its main anatomical subdivisions in living hominoids using

magnetic resonance ~ . Journal of Human Evolution, 38, 317-332. Senut, B., Pickford, M., Gommery, D., Mein, E, Cheboi, ~ & Coppens, Y. 2001. Fast hominid from the Miocene

(Lukeino Formation, Kenya). Comptes Rendus de l'Academie de Sciences de Paris, Series Ha (Sciences de la Terre et des plan~tes), 332, 137-144.

Slijper, E. J. 1958. Walvissen. Centen's Uitgevefij, Amsterdan~ Sponheimer, M. & Lee-Thorp, J. A. 1999. Isotopic evidence for the diet of an early hominid, Austra/op/thecus afr/-

canus. Science, 283, 368-370. Sprague, J. M. 1943. The hyoid bone region in placental mammals. American Journal of Anatomy, 72, 385-472. Tobias, E V. 1998. Water and Human Evolution. Out There, 35, 38-44. <http'J/allserv.rug.ac.be/,-mvaneech/out-

there&tm> Vaneechoutte, M. 2000. Water and human evolution. Human Evolution, 15, 243-251. <htlp'Jlallservzug.ac.bel,-mva-

neech/Symposittn~htrnl>

70 VEmt~GEN, MUNRO

Vaneechoutte, M. & Skoyles, J. 1999. The memetic origin of language: humans as musical primates. Journal of Memetics, 2. <htlp'J/allserva'ug.ac.te./ .-mvaneech/O~Jalml>

Verhaegen, M. 1992. Did robust aus~opithecines partly feed on hard parts of ~ e a e ? Human Evolution, 7, 63-64.

Vedaaegen, M. & Puech, P.-E 2000. Hominid lifestyle and diet r e c o n s i ~ paleo-environmental and comparative data Human Evolution, 15, 175-186.

Verhaegen, M., Puech, P.-E & Munro, S. 2002. Aquarbon~ ancestors? Trends in Ecology and Evolution, 17, 212- 217.

Walker, R. 1981. Diet and teeth - dietary hypotheses and human evolulioru Philosophical Transactions of the Royal Soc/ety London B, 292, 57-64.

Walter, R. 2000. Out ofAfr/ca. <http'JAvww.exrtca/hominids/outofafrica.cfin> Walt~ R. C., Buffier, R. T., Bruggemarm, J. H., Cafilhume, M. M. M., Berhe, S. M., Negassi, B., Libsekal, Y.,

Cheng, I-L, Edwards, R. L., von CoseL IL, N&audeau, D. & Gagnon, D. 2000. Early human occul~on of the Red Sea coast of Eritrea during the last interglacial. Nature, 405, 65-69.

Received" December 10, 2002 Accepted." June 10, 2003

Documents

Possible preadaptations to speech. A preliminary comparative