The earliest indications of the use of a symbolic notation to represent or visualize an astronomical cycle come from the Upper Palaeolithic period. At this time, in addition to striking cave art, engravings were made on small portable objects such as stones and pieces of bone and antler. Thousands of examples are known. A number of these are in the form of series of marks, and several were subjected to meticulous microscopic analysis by the American researcher Alexander Marshack. He concluded that the marks were not a simple decoration but carefully accumulated, often using different tools and techniques, over a period of time.
One famous example is part of an eagle’s wing discovered in a cave at Abri Blanchard in the Dordogne valley, France. Dated to around 30,000 b.c.e., it contains a series of notched marks in a serpentine pattern. Mar-shack proposed that these represent a tally of days. The assumption is that the earliest marks are those in the center of the pattern, and that marks were accumulated around existing ones. By following the line outwards and back and forth, we discover that there are about fifteen marks in each sweep before the direction changes.
The most prominent astronomical cycle is the phase cycle of the moon. In addition to being readily observable, it coincides with the female menstrual cycle. The moon’s phase cycle is certainly recognized among modern hunter-gatherer groups, although not necessarily universally (an apparent exception being Australian Aboriginals). The period of the lunar phase cycle (synodic month) is between twenty-nine and thirty days, so one interpretation of the Abri Blanchard bone is that it represents a tally in which the days of the waxing moon are marked off in one direction and those of the waning moon in the other; in other words it forms a rudimentary lunar calendar, maintained for about two months. Marshack suggested a similar lunar-calendar interpretation for patterns on a number of other Upper Palaeolithic portable artifacts.
Several criticisms can be made of the interpretation of the Abri Blan-chard bone as a lunar calendar. Two assumptions upon which it rests are that the number of marks between each “turn” in the line was significant and, second, that this represented the period between successive new or full moons. How easy might it be to fit other explanations? How can we judge a particular interpretation against the alternatives? At least one of the turns is not sharp, which gives greater flexibility in interpretation. Two of the lines could easily be interpreted as separate straight lines rather than part of the serpentine pattern. And there is the question of what exactly we mean by “new moon”: there is a one- or two-day period each month when the moon is not visible at all (astronomical new moon occurs in the middle of this), but it is the first reappearance of the crescent moon in the evening sky (the popular concept of “new moon”) that is the most significant event in visual terms, widely recognized even in the today’s world, from small indigenous groups to major religious calendars. Finally, although some of the marks appear round and others crescent-shaped, there is no apparent correlation between the shape of the marks themselves and the lunar phases.
All of these points introduce doubts in the interpretation of the Abri Blanchard bone as a lunar calendar. It is possible to address them by increasing the complexity of the explanation: for example, the left-to-right sequences contain more marks and can be taken to represent the days from full moon to new moon in the popular sense. But the potential for speculative argument seems endless. Indeed, the underlying microscopic evidence that the engravings represent tally marks or some other form of notation in the first place has been vigorously questioned.
And yet the underlying idea—that certain people in the Upper Palaeolithic would have recognized the phase cycle of the moon and may have attempted to record it—seems plausible enough. One approach would be to try to ascertain, hypothetically, how easy we would find it to “recognize” patterns in sets of markings that were in fact unintentional (for instance, caused by people sharpening tools) or had other meanings entirely. The conclusions could then be used as the basis for a formal statistical test. Yet even if lunar tallies or calendars were quite commonplace in the Upper Palaeolithic, they may not have been recorded at all consistently, in which case any attempt to identify sets of calendrical tallies recorded in a systematic way would be doomed to failure.
Studies of some of the many other Upper Palaeolithic engraved artifacts may clarify the issue. One example, another bone fragment known as the Tai plaque, has been interpreted by Marshack as a more sophisticated lunar calendar. And yet more complex designs give us greater flexibility in interpretation. This is not to say that such explanations are necessarily misguided, but rather that assessing them is no trivial matter. Identifying a methodology that will satisfy both scientists and social scientists is a challenge that has yet to be met by archaeologists.