British Archaeology, no 8, October 1995: Features

Before the introduction of farming, c 6,000 years ago, much of Britain was covered in forest. This forest was a major source of food and other materials for Mesolithic people; and it was this forest which the first farmers began to clear. It was presumably also from here that the great timbers of the earliest monuments were extracted. But what did the forest actually look like?

Until recently, archaeology has relied on the evidence of ancient pollen to answer that question. Pollen, however, can be hard to interpret. First, it gives ambiguous evidence about the proportions of different species of tree - some are under-represented, while others, such as pine, could be over-represented because their pollen can travel great distances. Second, pollen cannot tell us about the distribution of different trees. Take hazel, for example, whose nuts are one of the most frequent vegetable foods found on archaeological sites: was it scattered evenly through the forest, or concentrated in stands? Third, pollen can't tell us about the age structure of the forest.

Now, however, a team from Liverpool John Moores University (consisting of me, Dave Wilkinson and Alan Clapham) is trying to produce a clearer picture of what the ancient forest looked like. We are doing so by examining Britain's submerged forests, and our work is the first to attempt to use these macro-fossil remains - tree-stumps, leaves, seeds, and so on - to reconstruct the ecological structure of the natural prehistoric forest and its floor.

The initial results are encouraging. On the Lincolnshire coast north of Skegness, for instance, the presence of earthworm eggs showed that the trees, which were preserved in peat, had originally been growing not in peat but on a normal soil, for which there was no other evidence. By contrast, at Borth near Aberystwyth - one of the most famous submerged forest sites - the composition of the forest floor varies enormously, with peat (suggestive of open pools) alongside birch stumps and birch twigs, and elsewhere what appears to be a dryland floor.

Taken together, the results from a number of sites demonstrate what has long been suspected, that the late Mesolithic/early Neolithic forest was a kind of mosaic. Rather than being a homogenous mix of trees of different species, the forest on dryland consisted of stands of trees in which particular species were more common than others.

The possibility thus exists that prehistoric people used these stands selectively, for timber (as some trees are easier to clear than others), for food, and as a guide to where the best soils were for farming. This, in turn, could provide one potential explanation for the patchy distribution today of archaeological artefacts. Some parts of the forest, indeed, may have been truly impenetrable because of the existence of brambles.

Alongside this mosaic of dryland woods were others, such as those in valleys (for instance in the Lake District), fringed with alder, birch and willow, with birch and open pools further out onto the floodplain. Such forests must have looked very much like the present-day carr at Cors y Llyn, near Builth Wells in Powys, and this raises a fundamental issue for conservation. Whilst such modern woodlands may not be particularly old, they may actually have a character and function very similar to parts of the prehistoric, natural woodland. Should they be, therefore, protected as a specific type of `ancient woodland'?

Equally, the initial results of our work pose the question of whether designated ancient woodlands, like the tracts of coppiced oak and ash in parts of the Lake District, originated not in selective planting but in the selective retention of particular natural-woodland stands.

Two aspects of the initial results are directly relevant to archaeology. First, the size of the trees and their distribution shows that the wildwood was not wholly composed of massive trunks just waiting to be cut down and incorporated into monuments. Massive trees would have had to be selectively extracted from a large geographical area.

Second, it would have been easier for prehistoric people to cut down the majority of the smaller trees growing between the forest giants. If this occurred, then by the end of the Neolithic there could have been areas where larger trees were more numerous than smaller ones, producing a kind of `parkland' appearance. Such clearances would have produced irregular clearings. This prompts the question: could this kind of clearance-process be the origin of some of our small irregular field systems?

Tom Clare is a Senior Lecturer in Archaeology at John Moores University, Liverpool

Rome's navy was vital to the control of her empire, with roles in warfare, the protection of merchant shipping and the transportation of the emperor and his soldiers.

Yet our knowledge of the ships of this navy, and in particular their development, is severely limited. To date, much of our information has been drawn from pre- or post-Roman sources, often with the assumption that ship design remained more or less constant throughout the Roman period.

A recent study of the representations of military vessels on Roman coins, however, has shed new light on their evolution. The evidence suggests that a complete change in the function of the Roman navy took place over a period of three to four centuries. A navy based upon war fleets, intended solely for direct combat, evolved first into an organisation of multi-purpose military vessels and then into a number of river fleets, composed of small vessels intended almost exclusively for transport.

The earliest Roman coins with ships on them appear at the end of the 2nd century BC, and they continue down to the 3rd and 4th centuries AD. Over this period more than 250 coin types bear ships as their principal motif. This profusion of accurately dateable evidence is enormously valuable for refined dating of changes in ship design. The quantity of evidence also helps in overcoming some of the problems inherent in interpreting iconography. While ships on Roman coins cannot be taken as accurate copies of Roman vessels, features consistently attested on a range of coin types are likely to be accurate.

The evidence allows a much more detailed interpretation of the development of Roman military ships than any given previously, and suggests that significant alterations need to be made to the accepted chronology of changes in such craft. Two main periods of development are evident.

The first, in the latter half of the 1st century BC, is the introduction of a style of ship-building borrowed from Rome's Illyrian (or Balkan) enemies. The coins suggest that this new style is marked by the appearance of undecked vessels and the introduction of the forward-raking mast, the artemon, as seen on the coins of Mark Antony.

These developments indicate changes in the function of the Roman navy of the period. The ships of the Republic (from c 260BC to c 50BC) had been vessels intended primarily for combat, with full decks to provide a large area for mustering marines, a formidable ram, and an easily removable mast allowing a high degree of manoeuvrability. The 1st-century adoption of undecked vessels must have made military vessels less useful for direct combat, but more suitable for the portage of cargo and the carrying of men.

In line with this change, cabins now became the norm; and some ships have two (usually at the stern, but occasionally at the prow), suggesting an increased level of comfort - important if galleys were to be used for the transportation of dignitaries.

From the 2nd century AD, however, warships seem to be increasingly used for ferrying the legions; and a high proportion of craft now bear military standards.

The trend from specifically designed warships to general-purpose military vessels continues with a second major period of change in the late 3rd century AD. Ship types more traditionally associated with the Byzantine Empire now become prominent. These vessels have dominant fore- and stern-posts, carry fixed masts which would have made them less agile in combat, and are rowed from deck level. Some ships now have no ram, and are clearly intended exclusively for transport. It may be that ships previously associated with the northern seas had become prominent by this period, as many of the transport vessels are reminiscent of nothing so much as Viking longboats.

This change in the type of ships employed is likely to have been part of a general reorganisation of the fleet in the 3rd century. War fleets were apparently of little significance during the later Empire; they are seldom mentioned in contemporary documents, and increasing emphasis was placed on land-based defences, such as the Saxon Shore in Britain.

Evidence for river fleets does exist, however - the Notitia Dignitatum refers to barcae, boats, and these must be the small vessels depicted on many later Roman coins.

The profusion of coin types depicting ships declines dramatically between the 3rd and 4th centuries. Partly a consequence of a functional change in the coinage system, this also reflects the changing ideological nature of the Empire. The ship was an improbable symbol for what had become, by then, an inward-looking empire.

John Orna-Ornstein is a Curator of Roman Coins at the British Museum

Picture an ape which has been taught to manufacture and use stone tools. What is the image that comes to mind? A clever ape, or the very image of earliest Homo?

Most people would say they still see an ape. But, for years, apes and monkeys have been used for modelling early hominid society and intelligence, and over recent months a number of intriguing new ideas have been generated about early humans as a result of this type of research.

In one idea, the origins of human intelligence, tool-making and even language are placed many millions of years earlier than the archaeological record suggests; and in another, hominid society is seen as female-dominated and placid, in contrast to the `Man the Hunter' paradigm of early human behaviour.

Despite the intellectual difficulties of using modern primates as replicas of 2-4 million-year-old hominids, this type of research is nonetheless gaining intellectual ground, partly because recent fossil discoveries suggest that earliest Homo was - anatomically, at least - far closer to modern apes than to modern humans.

Australopithecus ramidus, for instance, the oldest known hominid discovered last year and dating from about 4.4 million years ago, had teeth and an overall skeleton that closely resembles that of the modern pygmy chimpanzee. Even Homo habilis, maker of the first stone tools some 2.5 million years ago, is now known to have had an ape-like posture with long arms and short legs, and a short growing-up period, quite unlike our extended childhood needed for the development of the human brain.

`Research into living primates is valuable, because so much of what we thought we knew about hominids is now being questioned,' said Chris Stringer of the Natural History Museum.

`Early hominid society is now being seen as much more ape-like, and less like us. Homo erectus [evolving c 2-1 million years ago] is the first hominid with a human (rather than ape-like) skeleton, and some people are now saying you can't project human models back further than about 50,000 years,' he said.

Primate research offers evidence of three main types. First, by demonstrating what apes can do (for instance, with tools), it provides a touchstone for judging the scale of human achievement seen in the archaeological record. Second, primate society provides possible models for hominid society and behaviour. And third, by demonstrating what apes cannot do, it gives possible indications of why we evolved into humans, and why apes - even when taught to make stone tools - remained apes.

Everyone knows that chimpanzees, our nearest primate relatives, use tools (mainly of vegetation) for various purposes in the wild. Some apes have also been taught to make stone tools in laboratory experiments. But it has generally been accepted that apes are unable (even when taught) to make as effective stone tools as the earliest-known hominid tools. These were made some 2.5 million years ago in Africa in the so-called Oldowan technological stage.

However, in a series of experiments reported in World Archaeology (June 1995), Dr Greg Westergaard of the National Institute of Health in Poolsville, USA, has demonstrated that capuchin monkeys in captivity spontaneously make stone tools - without being taught to do so - and use them to crack nuts, cut material and even deflesh bone.

Moreover, they make tools in the same four ways postulated for Oldowan hominids - holding two stones and hitting one with another, using a stone to hit another stone placed on a rock, striking a stone on a rock, and throwing a stone against a rock. Although the tools produced are still cruder than Oldowan tools, they are remarkably close. `It's a distinct possibility that capuchins actually make stone tools in the wild unobserved,' Dr Westergaard said.

In addition, and perhaps more surprisingly, Dr Westergaard has shown that in making tools capuchin monkeys are mostly right-handed. Right-handedness is assumed to have developed in hominids by 1.4-1.9 million years ago, and has been linked - through the development of hemisphere-specialisation in the human brain - to the origins of language.

What can be made of these findings? First, they suggest that some fairly late-developing human `achievements' were not distinctly human at all, but can be paralleled in the non-human primate world. Second, they point to the possibility that some artefacts taken to be hominid were in fact made by apes or monkeys.

And third, they suggest that the capacity for the development of higher intellectual faculties - a `neural substrate', as Dr Westergaard calls it - was present before the divergence of capuchin monkeys from the main primate lineage some 45 million years ago - many millions of years earlier than is now commonly accepted.

A similar claim has been made recently by Dr Dick Byrne, Reader in Psychology at St Andrews University, in The Thinking Ape, published by OUP earlier this year. Dr Byrne, who researches African primates in the field, argues in the book that the origins of `human intelligence' lie much further back than the development of language or the first stone tools, and stem from before the divergence of humans from other primates 9-5 million years ago.

Human intellect, he argues, is not dependent on language or complex social arrangements, but developed out of sophisticated strategies for subsistence, some of which are also employed by other primates (such as the use of tools to solve problems).

Although chimps generally only use tools made of wood, vine or plant stems, they gain enormous benefit from them, he explained. `This does seem to me to be a precursor of stone-tool manufacture. Our earliest ancestors probably also first used non-stone tools, but they simply haven't fossilized, and tool-use probably goes much further back than we'd think.'

Some of the most intriguing recent models of early hominid society have been produced through the study of the pygmy chimpanzee, or bonobo, whose skeleton is very similar to that of A ramidus, the earliest known hominid, and which shares more than 98 per cent of the genetic profile of modern humans.

One line of research, conducted at Georgia State University's Language Research Centre, touches on the origin of language, and suggests that the earliest humans may have been able, more or less, to speak. Bonobos, the research found, can modulate their voices far more than other apes and can `say' at least 300 words. One bonobo, known as Kanzi, has famously been taught the rudiments of American sign language.

More recently, bonobo research has suggested that the earliest humans may have been less male-dominated and warlike than has hitherto been supposed (mainly from aggressive, male-dominated chimpanzee models). Writing in Scientific American in March, Prof Frans de Waal of Emory University, USA, reported his own and other people's research that showed that bonobos - unlike chimpanzees - rarely fight among themselves, are dominated by females rather than males, and typically defuse tension by frequent and human-like use of non-reproductive sex - including missionary-position couplings, oral sex and genital-genital rubbing, particularly between females.

According to Prof de Waal, bonobos are likely to have undergone less transformation than either humans or chimpanzees - because of a marked ecological continuity in their habitat over millions of years - and so could most closely resemble the common ancestor of all three modern species, not only in anatomy but also in social behaviour.

Other researchers may not go so far. But according to Chris Stringer, it was `very, very interesting' that there were such marked differences between the two types of chimpanzee: `If you are trying to get a picture of our earliest ancestor you have to look at both types of chimp, as well as humans; and not only their anatomy, but their culture as well,' he said.

However, if early human society was ape-like, the question remains: how did later human society become human-like? Primate research can only scratch the surface of this vast question, but does so mainly by analysing what apes and monkeys cannot do. Writing in a recent Cambridge Archaeological Journal(October 1993), Dr Steve Mithen of Reading University makes two important points. First, apes seem unable to integrate tool-use, communication and intelligence, as humans can, but instead can only manage one pattern of thought at a time; and second, apes imitate each other less than humans, leading to less cumulative learning across generations.

Taking the matter further than that, however, is a problem. This is the great imponderable question of humanity, and solving it is a task that will no doubt keep us busy for generations to come.