Although whole pots are sometimes found included as gravegoods in burials, ceramics can break easily and archaeologists rarely find whole pots, abandoned where they were used at a site. It is much more common to find small bits of broken pot, called "potsherds," discarded in rubbish piles, or mixed in with other cultural debris. Some sites, such as the Hamdallahi site in Mali can contain thousands of potsherds in a thin layer of a single 1x1 meter excavation square.

How can archaeologists mak e sense out of such fragments? One step is to try to reconstruct whole pots from their pieces, using clues from the "morphology" (shape and size) and decorative style of the individual sherds. This is analogous to a 3-dimensional jigsaw puzzle. But how many pieces of a pot must an archaeologist find in order to determine the shape and size of the original pot? And which parts of the pot might be the most useful for reconstructing its original form?

Many archaeologists prefer to base their analysis of the ceramics at a site only on big sherds or reconstructed vessels (e.g. T. Huffman's analysis of the Gokomere/Ziwa ceramic tradition in southern Africa). On the other hand, if most pottery fragments recovered from a site are either too small or undiagnostic to attempt refitting, should they just be ignored? What can archaeologists learn about prehistoric peoples from the "attributes" (features) of their fragmentary pottery?

The Predynastic cemetery HK 43 at Hierakonpolis (Upper Egypt), which dates to 3600 BC, has the earliest recorded examples of hair extensions and hair die in the entire world.