The Form of the Argument

The aim of this section of the web site is to consider how one goes about making the argument for early seafaring on the basis of island archaeology. The task then is to review the logic and the inferences that are drawn when the archaeologist uses what is found on the islands in the Mediterranean Sea to make the case for early voyaging. Emphasis will be placed here on the large offshore islands where it is easier to show that voyages of some distance were made over the open sea. What is of interest for our present purposes is voyaging as an activity that was done on a repeated or regular basis (that is, when voyages to and from a given offshore island were undertaken with a frequency that falls somewhere in the range between one or two crossing per year and at least one crossing during the span of a human generation (that is, about 25 years in the remote past, which would mean just four voyages were made to the island over an arc of 100 years) and not as an activity that entailed rare events (including the possibility of accidental crossings). The latter would constitute a rather different matter from the one that we wish to study. In short, we are interested in learning more about how voyaging began as a way of life. Here attention will focus on the form of the argument and not on the details of individual case studies, which will be taken up in various papers at the meeting.

The basic argument can be summarized in the following terms: (1) the island that one sees on a map today was an island at the time of interest in the remote past, (2) when voyagers made their first trip to a given island, no one was living there, (3) early archaeological remains that are found on the island show that people reached it by voyaging over the sea (note that the first voyagers may have frequented the island only on an occasional and seasonal basis; they need not have occupied, settled or colonized it) and (4) there is no other way to explain the occurrence of the archaeological materials on the island. More will be said below about each of the steps in the argument. At the meeting, we can discuss whether there are other steps that should be added to this list.

Before we turn to the first step, it is worth saying a few words about the wider context in which the study of early voyaging in the Mediterranean is now situated. In many respects, the context in 2012 is quite different than the one just 20 years ago. Towards this end, it is worth mentioning briefly five points to set the stage today. Again, they will be taken up as topics for discussion at the meeting. (I) The earliest evidence for voyaging in the eastern Mediterranean now goes back to the time before the Neolithic period; previously, the study of this question traditionally focused on the Neolithic and the colonization of the various islands. What we are dealing with are hunter-gatherer or coastal foragers or else those who found themselves moving along the pathway toward agro-pastoralism but who still had some way to go before they became first farmers. (II) In the time before 13,000 years ago, there appears to be rather modest evidence for the presence of people on the large offshore islands in the Mediterranean. On the other hand, this may change as more fieldwork is carried out over the course of the next decade. (III) In the years before 11,000 years ago, there is still little or no good evidence for permanent forms of settlement (that is, occupation of an archaeological site on a year-round basis) on the offshore islands. Again, this may change in the years to come. In any event, there is the need for more studies of seasonality when it comes to pre-Neolithic sites, which occur on islands in the Mediterranean Sea. (IV) At the time we plan to focus on at the meeting (13,000 to 10,000 years ago), sea levels were lower (some 80 to 10 meters depending upon the time and also local tectonic movements), and they were in an on-going state of change. (V) Finally, we need to keep in mind the major changes in the earth’s climate, as they are now documented by deep cores made in the Greenland ice sheet, and, in particular, the Younger Dryas, the name of the cold snap that took place between ca. 12,800 and 11,600 years ago.

The First Step: The Inference that the Island was, in fact, an Island. One starts the argument by making the inference that the island where the archaeologist works today was an island at the time of interest in the past. One gives the reasons for why this is so. If the place were still attached to the mainland (in one way or another), then one would be in no position to make the case for voyaging. In addition, if the island separated from the mainland fairly recently (for example, a few thousand years before the time of interest), then the distance over the sea to the island would have been, in all likelihood, a comparatively short one. Accordingly, the case study itself will not support a claim for early voyaging over some distance. More will be said below about the case of short trips that were made in coastal waters.

The key factor at the first step is the bathymetry around an island. This, in combination with the sea-levels curves currently available for the last 18,000 years (see, for example, the one recently published in the first chapter of Submerged Prehistory and illustrated elsewhere on the web site) will give an initial sense of whether the island was an island or not at a given time. For example, if the bathymetric chart shows that the waters all around the island have depths of 200 meters or more, then it will be reasonable to think that it has been an island for a long run of time (sea level went down to a low point of ca. 120-130 meters below its modern level some 18,000 to 20,000 years ago). Thus, in the case of Cyprus where the waters all around the island have depths much greater than this (and deep bathymetry is observed widely on all sides of the island as well), one can infer with confidence that it has always been an offshore island for the last 20,000 years. While there may be some minor earth movements that have occurred locally (for example, the modest tectonic uplift observed on the west coast of Cyprus), they will not affect the big picture. Thus, islands such as Cyprus and Crete are good places to work for the archaeologist who wishes to make the case for early voyages over some distance.

On the other hand, there are islands in the Aegean Sea where it is much more difficult to say when an island actually formed. As we shall see at the meeting, this is the case of islands such as Lemnos and Gokceada. Without going into the details here, the challenge is that of working out the interaction between three different factors – bathymetry, trends in sea-level rise and local earth movements – in order to establish the time when the island began to separate from the mainland. Here it is worth commenting that each of the three factors is known today only to a certain level of approximation. In the case of sea-level rise, for instance, there are even different geo-physical models that one has to choose between (see the various articles by Lambeck, Peltier and Pirazzoli in the literature). As the study of island archaeology in the Mediterranean keeps moving back in time during the course of the 21st century, there will be the need to improve and refine what is known about all three of the factors.

The Second Step: The Inference that No One was on the Island in the Time before the early Voyagers first reached it. On the face of things, this is an inference that, for many years, appeared to be a rather simple and straightforward matter. As mentioned before, the conventional wisdom well into the 1990s was that the Mediterranean islands only began to be frequented in the Neolithic period. Hunter-gatherers were reluctant voyagers. Accordingly, the archaeologist assumed that no one was living on the offshore islands in the time before first farming. Of course, we have now moved beyond this assumption. There is, however, still a tendency for our thinking to remain in the shadow of this old idea. Indeed, once we begin to formulate the matter in terms of hunter and gathers and coastal foragers, the whole question of early voyaging becomes more complex and demanding. How far back in time on a given island can we really take the first appearance of pre-Neolithic sites? There is no simple answer to this question. It is something that we have to sort out on the basis of fieldwork on the island. In addition, there is an issue of methodology that arises here. One has to be careful and not confuse the current lack of evidence for something on a given island with evidence for its absence there. Due cautious is always called for when it comes to making inferences for absence in archaeology. In most case studies concerned with islands in the Mediterranean, the inference that human beings had not made their appearance on a given island before a specific time can only be made as a working hypothesis today.

A comment that needs to be made here is that the recovery and the documentation of a site on an offshore island dating back to 20,000 years ago do not necessarily mean that there is evidence for the continuous presence of hunter-gatherers on that island for the following 10,000 years. This too remains an open question until more fieldwork is done on the island. At this point, it is also worth adding that on islands such as Cyprus and Sardinia, where a certain amount of reconnaissance work has been done by those with Palaeolithic experience over the last decade or two, not much has come to light that dates the time before 13,000 years ago (on Cyprus, recall the results of the Amathos Survey as well as those of the paleontological excavation at the Agia Napa rock shelter). On the other hand, when the same archaeologists have conducted fieldwork in various regions on the mainland and also on islands that were once connected with the mainland down to the end of the Pleistocene, they have usually been able to recover scatters of chipped stone dating to the Upper Palaeolithic and the Middle Palaeolithic within a matter of few weeks and sometimes even in a few days. So there does seem to be something behind the shortage of Palaeolithic sites on the large offshore islands.

Finally, an added measure of caution is called for in the case of islands that formed only in more recent times (for instance, 16,000 to 10,000 years ago as a consequence of sea-level rise). Here the “new” or “young” island had a long history of being attached to the mainland, and at the time of its separation, it may well have witnessed the persistence of hunter-gatherers whose territory, for generations, had included this place. Thus, an island of this kind should be treated as a special case -- one where it may be difficult to make sound inferences at the second step in the argument. In any event, such a “new” island will probably not offer the opportunity to make an argument for voyaging over any real distance, as mentioned before.

The Third Step: The Inferences based on the Archaeological Remains that are found on the Island. At this step, one is dealing with positive evidence and its interpretation. The archaeologist would like the sites and archaeological materials found on a given island to have several positive features. First, the work in the field has led to the recovery of early sites at several different places on and around the island. Second, the excavations that have been conducted at early sites have produced archaeological remains that are found in situ. Third, a series of good radiocarbon dates – ideally AMS dates on organic materials that derive from plants and animals with short lives – is available for each of the excavated sites. Of course, this is a wish list. Normally, one is working by steps of approximation on all three fronts. A given island at the present time may have only one or two pre-Neolithic sites on it and only one where an excavation has been conducted so far. Today the archaeologist is making inferences at the start of a long learning curve. It is too soon to regard the inferences that we are making as definitive ones. At the same time, what is coming to light on other islands and the inferences that are being made by the archaeologists who are working there will influence the claims that the archaeologist can make in his or her own case study. In short, the endeavor that the archaeologist engages in at this step in the argument has to be seen as a dynamic and open-ended one.

This is a good place to consider two quite different models of early sea going. They may lead to a better understanding of what is involved in making the argument for early voyaging -- again in the sense of voyages on the open sea that covered a certain distance. On one hand, there is the model of the “young” island that has separated recently (in the geological sense of time) from the mainland. At the time of separation, there would have been a short distance between the mainland and the new island, and it would have been possible to move in coastal waters between the two even in a small boat that was quite simple in its form and construction. For the hunter-gatherer or the coastal forager who made such a trip, the new island may well have formed part of the territory that one’s family and ancestors had known and exploited for years (that is, when it was still attached to the mainland). Crossing over to such an island did not represent going out to a new place. Instead, what it meant was staying in touch with one’s “home” territory. In subsequent generations, as sea level continued to rise, the distance between the island and the mainland would have slowly kept getting larger. And sooner or later, as the distance increased, one would have become more proficient in moving across coastal waters. However, even after many generations, such an approach to going to sea need not have involved long-distance voyaging. The island of Lemnos in the northern Aegean would make a good case study here.

On the other hand, there is a quite different model: that of going out to an offshore island, a remote and unknown place, which required moving over a much greater distance. In order to make successful trips between the mainland and the island, it would make good sense to build a better boat and to have skills and experience in crossing the open sea. In short, this is the model of voyaging, and Cyprus offers a good case study. In effect, these two models represent the two ends of a spectrum; one can think of other models of early sea-going that fall at various points between them. In more general terms, it is reasonable to think that the first model was the older of the two. In all likelihood, it constitutes the point of departure for the whole story that we would like to investigate.

The Fourth Step: The Inference that there is no other Way to explain the Occurrence of the Archaeological Material on the Island. At the same Time, in making the Argument for early Voyaging, there is often the tacit Assumption that the Archaeological Remains, as they are currently known, tell the full Story. The task at this step in the argument is to make a concerted effort to think of other ways to explain the archaeological material and to have good reasons for excluding them. In short, one puts on the shoes of the devil’s advocate. No attempt will be made here to discuss this step in the argument at length. Much will depend on the details of a given case study. To begin with, one has to consider the possibility that the archaeological material derives from a rare event -- one that even involved chance. For example, there is the possible case of accidental rafting that the archaeologist has to consider. In order to exclude it, one will have to focus on those patterns in the archaeological record, which show that early voyaging was intentional in character. In addition, one may have to return to the second step in the argument and make sure that the island really was an island: that is, one has to develop an even stronger geological case for the island not being connected with the mainland at the time of interest.

There is still another complication that comes into play at the fourth step in the argument for early voyaging. Is the archaeologist in a given case study really in a good position to make inferences that are based on all of the main or important elements in the story of early voyaging? Or are some of the key pieces in the puzzle still missing? An example may help to illustrate the nature of the problem. There is, one can argue, a good chance that the archaeologist will make the inference that the time when voyagers first began to arrive on a given island is too recent. Indeed, the date of the time of first arrival may be off by several thousand years. Here the problem stems from the fact that so little is known about the submerged prehistory of the large offshore islands in the Mediterranean. If one goes back to the time of 12,000 years ago, one is dealing with sea levels that were on the order of 60 to 70 meters lower and a shoreline that was some 1 to 2 kilometers wider in many places than it is today. Now if we are prepared to go back even further in time to say 15,000 years ago, the respective numbers will become even deeper and wider. In other words, there is a large submerged space – one involving just those places where early coastal foragers probably kept their boats and made their campsites – that has yet to be explored at all. A key element in the story is definitely missing. Thus, in thinking about the form of the argument for early voyaging, there are still serious limitations with regard to the places where we as archaeologists has conducted our fieldwork so far. In short, we return to the issues that were raised in the section above called “Getting the Boat Moving.”