There are two basic kinds of chronological approach.  Ideally we use both and cross-reference them to produce an overall chronology.

 

In a stratigraphic sequence, the series of events clearly happened at different times.  From this we know the order of events, but not the extent of the time involved in each.  A series of deposits can sometimes take centuries to accumulate, but in other cases it may all happen in a few minutes.  Artefacts found in the layers can help us to decide the time-intervals within a sequence.  The presence of datable objects in recognisable typologies such as coins and pottery give us a good indication of the length of time involved in each part of the sequence (this is known as dating by association).  Relative chronologies were all archaeologists had to depend on until the mid-20th century, and they are still an essential part of understanding archaeology.  However, they can only give us a generalised time-frame which is referenced to its own elements and to similar sequences observed at other sites.  It does not have any direct attachment to an independent scientific means of measuring time.  For this we need to turn to absolute chronology.

 

Absolute Chronology is dependent on a direct link between the archaeology and an independent means of measuring time in years. Except in cases where inscriptions or documentary evidence provide this link, absolute chronology is only possible through the use of scientific techniques which fix evidence on to years before the present (BP = equivalent to years before AD 1950) – used more commonly in techniques for very long chronologies, or on to standard calendar years measured in BC and AD.  Even so, it is less common for an absolute technique to give us a single year, month or even time of day (dendrochronology is the best for this, but it is only feasible where enough of the right parts of a tree are preserved).  Other more common techniques, such as radiocarbon, give us a number of years during which the event we are dating most probably happened. This is known as a Standard Deviation.