Jason Edmonds, Cosumnes River College
Styles come and go over time. A new style is invented and, if it catches on, rises in popularity. At some point, its popularity peaks and then it fades away. As old styles recede, new styles emerge to replace them and begin the cycle again. This is true for cars, clothing, music, and memes and for ancient pottery. American archaeologists working in the early twentieth century combined this observation with data collected from excavations of stratified sites to create the relative dating technique called seriation.
Relative dating determines a chronological order without reference to calendrical dates. It is used to reconstruct the order of events and cannot directly determine the amount of time that has passed between individual events.
Seriation was frequently used by archaeologists working in the classificatory-historical paradigm during the first half of the twentieth century since they were interested in reconstructing cultural histories in relation to time and space. And reconstructing cultural histories continues to be an important part of current analyses of the past. It lays the foundation for further research questions regarding function, process, change, and explanation. Seriation is an effective tool for constructing cultural chronologies and defining culture areas. In this exercise, you will use frequency seriation to reconstruct a culture’s history.
For the purposes of this assignment, each assemblage represents surface collections from an archaeological site. Your goal is to determine the periods of popularity for each pottery style and then the temporal sequence of the pottery styles.
1. The following data set of archaeological ceramics provides the number of pieces of each type of pottery found in six assemblages.
First, compute the relative frequency of each pottery style present in the assemblages (rows). For example, assemblage 3 contains 98 pot sherds and 10 of the sherds are corrugated. To determine the relative frequency of the corrugated sherds, you divide 10 by 98 (10/98 = 0.102 = 10%). Round the frequencies up or down following the usual convention (round 5–9 up and 1–4 down). Write the frequencies in the space provided to the right of the numbers in the cells.
| |
Type 1 |
Type 2 |
Type 3 |
Type 4 |
Type 5 |
|
| |
Corrugated |
Black on white |
Plain (buff) |
Plain (red) |
Black on red |
Total |
| |
# |
% |
# |
% |
# |
% |
# |
% |
# |
% |
|
| Assemblage 1 |
30 |
|
0 |
|
89 |
|
8 |
|
0 |
|
127 |
| Assemblage 2 |
31 |
|
27 |
|
103 |
|
25 |
|
21 |
|
207 |
| Assemblage 3 |
10 |
|
0 |
|
88 |
|
0 |
|
0 |
|
98 |
| Assemblage 4 |
8 |
|
33 |
|
38 |
|
4 |
|
47 |
|
130 |
| Assemblage 5 |
34 |
|
10 |
|
119 |
|
30 |
|
6 |
|
199 |
| Assemblage 6 |
13 |
|
20 |
|
47 |
|
8 |
|
22 |
|
110 |
2. Next, transfer your frequency data to the following graph, which is divided by solid and dashed lines, by shading in cells. Each cell between dashed lines represents 10%. So, to transfer a frequency of 40% for a particular pottery style, you would shade in 2 segments to the left and 2 segments to the right of the solid line under Type. Frequencies of 0% require no shading. Shade the cells for all six assemblages, making sure to keep each assemblage together in a single row and to label the row with the assemblage number.
Type 1 Type 2 Type 3 Type 4 Type 5
50% 0% 50% 0% 50% 0% 50% 0% 50% 0% 50%
3. Create a seriation from your data by moving the assemblages up and down to put them in the correct order by frequency.
Simply filling in the graph is not creating the seriation. You have to move the assemblages around until you achieve the correct order by creating the “battleship curves” typical of frequency seriation. The simplest way to achieve this is to cut each row of the chart into a strip and rearrange the strips until you arrive at the correct order. Be sure to keep the frequencies of each row together.
Following is an example and some hints to help you with the process.
Example
Frequency seriation showing percentages of artifact types by stratigraphic levels.
| |
Type A |
Type B |
Type C |
Type D |
| Level 1 |
80% |
0% |
0% |
20% |
| Level 2 |
50% |
0% |
10% |
40% |
| Level 3 |
20% |
10% |
20% |
50% |
| Level 4 |
0% |
30% |
50% |
20% |
| Level 5 |
0% |
70% |
20% |
10% |
This is a perfect seriation since it shows the expected correlation between the stratigraphic levels at a single site. Having a stratigraphic sequence can indicate whether a seriation is correct.
Helpful Hints
- Notice that all of the zero percents are located at the top or bottom of the order. Zero percents are not allowed to appear in the middle of a sequence for an artifact style. So try moving the assemblages containing many zeros to the top or bottom of your order.
- Each artifact style can have only one peak in popularity. The percentage of the peak will vary, but the frequency of a style cannot bounce up and down in a correct seriation. In the example, Types C and D peak at 50%. Above and below the 50%, the frequencies decrease. They do not rise and fall again. Your seriation should follow this pattern.
- Also notice that some of the styles in the example do not create complete battleship curves. Type D comes closest to a full curve; in others, there are only the beginning or end of a curve, which is okay. Your seriation will likely be similar. As long as you follow the other hints, partial curves are not a problem.
Questions
Answer the following questions.
- What assumptions are necessary for seriation to work?
- What does the “battleship” curve represent?
- Can you determine which direction time is going (i.e., which sample is the youngest)?
- What are some potential problems with seriation?
- What are some benefits of seriation over other dating techniques?
- What other methods or techniques can be used to confirm or verify a seriation?
