The evolution of walking (bipedalism)



We are seen as the only great ape to have stood upright and taken strides forward.

Why this event took place is surrounded by a lot of hypotheses.  These include

1. savannah theory: leaving the trees to live on the ground (Dart, 1953)

2. brachiation hypothesis: it evolved from walking in trees from branch to branch (Gabo, 1996)

3. pre-common-ancestor hypothesis: that our common ancestor with chimpanzees was bipedal (author’s own consideration)

4. rock climbing hypothesis at the African rift valley (Winder et al., 2013)

5. The controversial Aquatic Ape hypothesis: that it evolved from wading in lakes and rivers (Morgan, 1997).

So in this article we’ll be looking at

a) how other species walk upright,

b) why the human skeleton is specialised for walking, the problems that causes us, and

c) why it might have evolved covering the 5 points above.

First though, a little background to help:

Bipedalism (where the skeleton has become dedicated to it) and tool-making are traditionally seen as the first defining marks of humans in the archaeological record.  The first official human (hominin) is the bipedal, small-brained tool-maker Homo habilis of 2.5 million years ago.  For a long time, the first bipedal ape (hominid) was accepted to be Lucy, the Pliocene Australopithecus afarensis 3.6 million years ago.  Hominid refers to all non-human apes in this article.  Hominin refers to all human apes.

Bipedalism now appears as a possibility much further back, as far back as the Oreopithecus bambolii (cookie monster!) upright Greek ape of 9 million years ago (Wayman, 2011).  The Oreopithecus may actually be very important in understanding bipedalism, as we’ll see later.   However, most of the hypotheses in the article are based on Lucy’s bipedalism

Walking upright in other species

images (4)

This gorilla appears to be female and is wading through a shallow river. orang-utans also wade in this fashion and both may use branches to gauge depth.

Humans are not unique in their bipedalism.  Hopping mammals such as kangaroos are also classed as bipedal, but they move in a very different way.  Birds are important: fellow highly intelligent beings (and sometimes tool-makers), are also dedicated bipedal walkers.

The Great Apes have their upright walking moments too but their skeletons are not designed for bipedalism.  Orang-utans do the most upright walking of all, which they use in the trees, walking along branches while holding upper branches.  Like gorillas, they also walk upright while wading.

Bonobos (pygmy chimpanzees) are the next great walkers.  These are a more docile and sociable species of chimpanzee that are matriarchal and peaceful natured than common chimpanzees.  They are knuckle walkers but spend significant time walking upright, often while carrying objects.  They can look very human when walking but their pelvis is not designed for bipedalism.

Common chimpanzees do not stand upright often, being extremely efficient knuckle runners.  Standing upright is for display and for performing tasks that require reaching, though captive humanised common chimpanzees will stand upright more.

Gorillas do not walk upright very much at all, but it is seen when they are wading through rivers, often holding a long branch for measuring depth.  The young walk upright most frequently and the adult males sometimes rise upright in display.

Of the lesser apes, gibbons are also known for walking upright, which they do on the group, with their long arms held up in the air.  Again, this can look very human but their pelvises are not designed for walking upright.

Our very different physical and skeleton makes researchers wonder what was different in our evolution.  Only we, of the Great Apes, have the pelvis and leg morphology of upright walkers.

Homo sapien bipedalism: the skeleton


Human and common chimpanzee, though with great similarities have very different locomotor features.  Note the legs, feet, pelvis, spine, and skull angle

Our skeleton has evolved for walking on two legs (bipedalism) with our curved spine, flat feet, hip and knee rotations and narrowed pelvis.  It’s actually a strange adaptation in some ways: we are less stable on our feet than quadrupeds and a narrowing of the pelvis is very strange in a species whose brains have evolved to be very large.  It would make much better sense for a large brained species to be on all fours, with a wide pelvis for childbirth.  This makes us wonder if the selection for walking in humans was very powerful.  When brains did start to really grow big, in the Homo erectus, 1.8 million years ago, the pelvis was getting narrower.  It’s a strange contradiction.

Why did bipedalism evolve?


  1.  Savannah hypothesis (Dart, 1953)

Pioneered in the 1960s, savannah theory has now fallen out of fashion.  It was a simple hypothesis: that Africa changed abruptly from dense forest to savannah at the start of the Pliocene.  Once living on the ground, humans became bipedal because of a need to carry food and stone tools, to hunt, and to look over the grass for predators.

Its downfall is very simple.  We now know that the dense forests of Africa did not change abruptly into a savannah, but into open woodlands (mosaic habitat) with lakes and rivers also.  Our australopithecine ancestors were living in trees, not on a savannah, and their long arms for moving in trees are additional evidence.

2.   Brachiation mosaic hypothesis (Gabo, 1996)

This leads naturally on from our improved understanding of the ecology of the rift valley.  We can see from Lucy’s skeleton that she was bipedal, but she was also specialised for swinging in trees.  So it is likely she made full use of her mosaic habitat.

Orang-utans walk in trees as they travel high up in forest canopies, also using their arms, so it suggested that this manner of locomotion was how bipedalism was selected for, as this hominid combined walking on the ground with use of hands for carrying and for into and rushing through trees.

One potential problem with this is the possibility that bipedalism may have evolved long before the change to a mosaic environment.

      3.   Pre-common ancestor hypothesis (author’s own thoughts)


Genetics currently suggest that the split with our common ancestor with chimpanzees took place 7-9 million years ago.

Current dating of Hominid fossils places the bipedal Greek Oreopithecus bambolii of 9 million years ago in the time point (but not at the African Rift Valley where hominin evolution seems to have taken place.  However the Sahelanthropus tchadensis of Chad has been diagnosed as bipedal and is dated at 7 million years ago (Su, 2013)

This opens the possibility that bipedalism was widespread in hominids before the common ancestor split and therefore took place in dense forest.

It also raises the possibility that chimpanzees were once bipedal and evolved knuckle walking as an adaptation to forest floor living (dominant male communities have an evolutionary drive towards greater size in a species, driving apes from trees and to the forest floor, where knuckle walking is the most efficient locomotion).  This is again my own hypothesis.

Potential problems are dating errors: if these early walkers have been dated too old or if the split occurred far earlier than currently believed.

      4.   Rock-climbing hypothesis (Winder et al. 2013)

An important new hypothesis appeared in Antiquity journal this year that suggests that climbing may have been important in early human evolution and this takes us back to the African Rift Valley at the start of the Pliocene.  Here early hominins would have been negotiating a rocky terrain with elevations, using their hands, arms and legs in new ways.  This terrain was being generated by the great rift that was, and still is, active here.

Therefore we have hominids climbing, reaching up with their hands, changing their ankle flexion, as they find water, refuge and protein sources on the rocky outcrops.  This would have combined with use of woodland areas and travelling across open ground.

It makes good sense and again, the only difficulty is dating.  Was the evolution of bipedalism before or after the split with chimpanzees?

     5.  Aquatic Ape Hypothesis (Morgan, 1997)

I’ve left the most controversial until last.  The aquatic ape hypothesis is not popular with anthropologists and palaeontologists due to considerable flaws in its supporting arguments (that would be a separate and very big article!).  It was a wonderfully researched book and did offer a better explanation of bipedalism than Savannah theory, but it has dated itself with its evidence (from human morphology interpreted as aquatic in nature) being now known to be erroneous.

However, there is one possibility in there to be gleaned from it and that is the value of wading in the selection for bipedalism.   Orang-utans and gorillas both travel through water and wade to do so with their arm held upwards.  The Pliocene African Rift Valley had many more rivers and lakes than it does now, and in fact, hominin fossils can accumulate around these extinct lakes and in water deposits.  Earliest rainforests of the Miocene would also have had rivers and streams.

So I have taken this from Morgan’s theory as a possibility suitable for serious consideration.


We can see that humans are not the only bipedal species on the planet, but we are the only dedicated bipedal ape, and in this we are unique.  Therefore there must have been an evolutionary selection factor.

The question though is when and where?  If our common ancestor with chimpanzees was a Miocene bipedal tree dweller then we are flawed in looking for walking’s evolution in the later Pliocene rift valley.  Indeed if arboreal (tree) bipedalism was present in Greece at this time, it may have been a widespread hominid trait during the Miocene.

If it did evolve during the Miocene then it could be that walking in trees and wading were the factors.

If it evolved in the African Pliocene then it is likely that tree walking, rock climbing, wading and walking on open land combined with use of the hands for carrying and tools were factors in its evolution.

Whatever the reasons for our retained and refined bipedalism, they must have been very strong ones, because they defied the difficulties of childbirth and increase in brain-size.  It remains a very tantalising evolutionary question.



Dart R (1953) ‘The predatory transition from ape to man’ reprinted in (1977) R Leakey and R Lewin (eds.) Origins London:  Macdonald and Jane’s

Gabo D L (1996) ‘Climbing, brachiation, and terrestrial quadrupedalism: Historical precursors of hominid bipedalism’, American Journal of Physical Anthropology 101 (1), 55-92

Morgan E (1997) The aquatic ape hypothesis London:  Souvenir Press Ltd

Su D F (2013) ‘The earliest hominins: sahelanthropus, orrorin and ardipithecus’ Nature Education Knowledge [online] available at accessed 24/09/2013

Wayman E (2011) ‘Hominin Hunting: human evolution’s cookie monster, oreopithecus’ [online] at accessed 24/19/2013

Winder I C, King G C P. Deves M and Bailey G N (2013) ‘Complex topography and human evolution: the missing link’, Antiquity  87 (336), 333-349



  1. The evolution of human bipedalism is not so tantalising, if we only analyse it into smaller elements & use the comparative data, eg,
    -which animals have very long & stretched legs?
    -which animals have the head-body-legs in 1 line?
    -which animals tend to have vertical spines?
    The answers are resp.: wading spp, diving spp, arboreal spp.
    This suggests that we evolved from vertical (below-branch) climbers to diving & wading creatures.
    Google, eg,
    -Greg Laden blog Verhaegen
    -new directions in palaeoanthropology Verhaegen
    -econiche Homo

    In any case, open plain or savannah ‘explanations’ don’t explain anything:
    “Bipedalism — It is often stated that human locomotion was an adaptation to running on the open plains, which is illustrated by expressions such as ‘Savannahstan’, ‘endurance running’, ‘born to run’, ‘le singe coureur’ etc., even on the cover of the most influential scientific journals. Verhaegen et al.(2007) disproved in detail all endurance running arguments (Bramble & Lieberman 2004) that our Homo ancestors during most of the Pleistocene were adapted to running over open plains. When we analyse human locomotion into more elementary components, the running ‘explanation’ appears to be a just-so interpretation (cherry-picking): Bramble & Lieberman (2004) interpret every locomotor trait in humans as having evolved ‘for’ running, without even considering possible wading or swimming scenarios. A comparative approach shows that, for each trait, semi-aquatic scenarios provide more parsimonious explanations (google ‘econiche Homo’ table 4), and that extant human running is a secondary and conspicuously imperfect adaptation which evolved late in the human past, e.g. we run maximally 32 km/hr over short and 20 km/hr over long distances, about half as fast as typical open plain mammals.” (Hum.Evol.28:237-266, 2013).

    Human Evolution now publishes in 2 special editions the proceedings of the symposium on human waterside evolution ‘Human Evolution: Past, Present & Future’ in London 8-10 May 2013:
    SPECIAL EDITION PART 1 (end 2013)
    – Peter Rhys-Evans: Introduction
    – Stephen Oppenheimer: Human’s Association with Water Bodies: the ‘Exaggerated Diving Reflex’ and its Relationship with the Evolutionary Allometry of Human Pelvic and Brain Sizes
    – JH Langdon: Human Ecological Breadth: Why Neither Savanna nor Aquatic Hypotheses can Hold Water
    – Stephen Munro: Endurance Running versus Underwater Foraging: an Anatomical and Palaeoecological Perspective
    – Algis Kuliukas: Wading Hypotheses of the Origin of Human Bipedalism
    – Marc Verhaegen: The Aquatic Ape Evolves: Common Misconceptions and Unproven Assumptions about the So-Called Aquatic Ape Hypothesis
    – CL Broadhurst & Michael Crawford: The Epigenetic Emergence of Culture at the Coastline: Interaction of Genes, Nutrition, Environment and Demography
    SPECIAL EDITION PART 2 (begin 2014) with 12 contributions

  2. An update of the littoral theory (e.g. my 2013 paper Hum.Evol.28:237-266). The conventional paleo-anthropological view – that human ancestors became bipedal by moving from the forests to the plains (schematically: ape=>human = forest=>plain = quadru-=>bipedal) – is biologically & physiologically unlikely, e.g. primates that move from forest to plain become more, not less quadrupedal (“the baboon paradox”); sweating requires salt & water (both scarce in arid grasslands); etc. An account of human evolution which mentions possible arboreal & terrestrial but not shallow-aquatic milieus is incomplete.
    Comparative, paleo-environmental & other data show:
    (1) Plio-Pleistocene australopithecines were typically wetland species (K.Reed 1997). This helps explain the remarkable combination of bipedality (e.g. for wading) & curved hand-bones (for climbing arms overhead). Human fetuses never have hand-like feet, but prenatal African apes have more humanlike feet (with longer & adducted big toes) which later become more hand-like (C.Coon 1954). This suggests Pan & Gorilla had more bipedal ancestors (e.g. parttime wading for papyrus, frogbit, waterlilies etc.), google e.g. bonobo wading, or gorilla bai.
    (2) Our Pleistocene ancestors (archaic Homo) did not disperse intercontinentally walking or running over the open grasslands, but followed African & Eurasian coasts & rivers (coastal dispersal model, S.Munro 2010), walking & wading bipedally & parttime diving for waterside, littoral & shallow-aquatic foods (which are richest in brain-specific nutrients such as DHA, e.g. S.Cunnane 2005), even colonizing islands overseas: Flores, Crete, Cyprus etc.
    Homo’s diet included animal (e.g. shellfish to be opened with hard tools, waterside carcasses of herbivores & marine mammals, salmon & other fish) as well as plant foods (e.g. traces of waterlily roots in neanderthal dental calculus & of cattails on their tools).
    Homo’s brain enlargement (e.g. DHA) & parttime shallow diving (which requires breathing control) were preadaptive to human spoken language.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: