Necessary for basic description of what was found (often before you know anything else)
For artifacts: Morphological types
Oval-shaped bifacial point, triangular point, etc.
Tumilaca style ceramic sherd; sherd with two or more paint colors; etc.
For plants or animals: Species, genus, bird vs. large mammal vs. small mammal vs. reptile; body or plant part (femur, seed, leaf, etc.)
Why do we have to categorize things?
that is, why do we need typologies?
because they allow us to see patterns in masses of data
say you have 1000 whole and broken points from each of 3 layers at each of 5 sites
you can't look at them all at once
even if you could, you would be overwhelmed by all the variation in size, details of the shapes, material, etc.,
but if you sort them into piles of similar types, then you start to see the patterns:
say the lowest level at site A has 20% triangular points, the middle level has 40% triangular points, and the top has 60% triangular points
now you have a pattern you can begin to think about
There are usually many ways of categorizing any collection of things
different categorizations may reflect different aspects of the same artifacts
example with lithic (stone) points (arrowheads, spearheads, knife blades, etc.)
we could categorize the points by forms (shapes)
which might be used for different functions
so hunting camps might have different forms of points than permanent settlements of the same culture at the same time
we could categorize the points by the material (kind of stone) used
which might change over time as relations with neighboring groups that control the quarries change
so red chert points might be earlier and black obsidian points might be later
regardless of form
we could categorize the points by size
size of certain forms might vary with economic status of the family that owned them
maybe better-off families could have larger stone knives, while poorer families had to make do with the smaller, resharpened dregs of the same form and material
but in another region or time period, these same categories might reflect completely different causes
form might change over time with shifting fashion, without affecting the function
material might reflect socioeconomic status if some materials were more rare and prized than others
size, rather than shape, might be the most important variable for determining function
there is often no way to know in advance what the significance of any categorization scheme is
you just have to try it out and see what the distribution of the types in time and space suggests
often you start with a hypothesis that justifies a certain typology
frequently based on ethnographic analogy (hints from living people)
say, different shapes of points might be made by different ethnic groups, so you choose to use form categories
or the angle of the cutting edge might be related to the use of the tool, so you record that
but you don't really know in advance if the categorization is going to tell you anything
each categorization might be useful for a different kind of analysis
no one categorization is necessarily more correct than another
as long as it is well enough defined that people can agree on which objects are which type
different classifications just have different uses for archaeologists
or some may prove to have no particular use at all - too bad.
most categorization schemes are by definition "etic"
they are constructed by archaeologists and don't necessarily match any categories that the original makers of the objects employed
some people try to reconstruct these "emic" categories
but that is a very hard thing to do
how could you tell what types people had in mind when they made them?
or "hunting points", "showing-off points", "points for use in the house"
or "points in my family's style", "points in Sam's family's style", "points in Eva's family's style"
there may be some subtle ways to get hints of the "emic" categories by looking at their distribution within and between sites, but it won't be easy
and not necessary for many purposes
the "etic" categories might be fine for distinguishing time periods, different activities in different sites, etc.
So, what do you actually do when faced with boxes full of artifacts?
invent a typology that seems like it might be useful for the question you are interested in
say, small, medium, and large points
or you might record this as a length measurement for each
classify the artifacts according to that typology and record the numbers, weight, or whatever of each type that were found in each context
that might be strata within each excavated unit at a site
or different areas of a site if the artifacts were collected from specified locations on the surface
or different sites
then you look for patterns in the distribution of the types in space and time
hoping to find patterns that suggest something interesting, or confirm some hypothesis
like "point size increased from the lowest (earliest) strata to the highest (latest) ones"
then you suggest an interpretation of the pattern, or proclaim your hypothesis supported or not
"the increase in point sizes supports my hypothesis that these people shifted from hunting birds to hunting mammoths"
someone else with a different question might very well have to go through all the artifacts all over again
someone interested in these people's exchange relationships with their neighbors might want to know about the materials they used for their points
but you didn't collect that information with your size categories
so they have to go and categorize the artifacts a different way, record that information, and do their own analysis
so why not just record all the variables at once and be done with it?
there is no practical way to record all the useful information in one typology
because there are far too many possibly interesting variables
you just can't know what someone might think is relevant for some future study
which might require special knowledge of geology, chemistry, specialized measurements, etc.
and recording a lot of information about each object would be so laborious that you might not make it through enough of the collection to be able to see the patterns at all
if you record 50 measurements per point, but only finish recording the points from one stratum, you will never detect the change in size from the bottom to the top of the excavation
so you have to balance how comprehensive your typology is versus how practical it is to use
it has to record enough variables (say, general form, material, a measurement or two) to let you detect the patterns you are interested in
while not recording so many that you spend your whole budget and lab season documenting the contents of the first box in the warehouse
I personally tend to err on the side of collecting too much detail
so I have computer databases with a lot of detail about what was found in every stratum of some of my excavations
for example, not only the amount of corn, but the amount of corn cobs, vs. kernels, vs. husks, vs. stalks, vs. flowers
each of which may reflect different activities
There is so much interesting variation in this data that it will probably never be fully exploited
I can keep going back to the computer databases to answer questions I had not even imagined when I did the field and lab work
but this approach is so slow that I also have excavated collections that I still have not categorized and recorded
(beyond a basic inventory of plant material, ceramics, etc. per stratum, which is required for keeping track of storage and planning future lab work)
so this approach gives me lots of detail for some sites, but has prevented me from having almost any detail at all about others
that limits my ability to contrast different areas or cultures, for example
and in spite of this, there is a lot of data that my detailed typologies have not captured
like paste types of ceramics, which can be very useful indicators of change over time and exchange among different groups
some day, someone may go through my collections, recategorize all the ceramics by paste types, and come up with some interesting conclusions
Mike Moseley's advice: Never collect more data than you can publish
a bit harsh, but worth keeping in mind
the down side: since you don't know in advance what data you will need or find useful, if you use types that are too simple, they may not tell you much
For controlling time (Thomas's "temporal types")
I disagree with Thomas's phrasing here; these are not really different kinds of types
They are simply types that are found to be restricted to certain strata or time periods
It could turn out that white points are earlier than red points
Or it could turn out that oval points are earlier than triangular points
Or maybe both are present throughout the whole sequence
In which case they are not temporal types
Seriation
This term is sloppily used to refer to two different things:
a documented pattern of change over time
and a method for discovering that pattern
Documented pattern: the discovery and presentation of a pattern of changing popularity of different types over time, like the changing prevalence of different lighting technologies in the text (p. 107)
often portrayed as "battleship curves" (p. 107)
often based on stratigraphy
imagine that each 10-year slice of the chart on p. 107 is a stratum at a dump, in which parts of lamps were found
Method: a method of creating these battleship curves based solely on the artifacts themselves, or sets of artifacts, with little or no help from stratigraphy
used when no well-stratified deposits are available
places where sites were often occupied for only one period
cemeteries where burials of different periods are scattered around but are not stacked on top of each other (no stratigraphic relationships to indicate order)
etc.
based on the idea that variation in characteristics of artifacts, or the relative popularity of certain types, often varies over time in a battleship curve fashion
say you excavate a bunch of historical dumps
each used for a short period
you record the proportions of fragments of different types of lamps in each dump
then you try to arrange the dumps in order so that they form a battleship pattern of gradual change
or you look at a bunch of decorated ceramic pots
you record many traits on each one
then you try to put them in an order such that each trait is used for a continuous period
say, the earliest pots are black and round
later ones are still black, but have a rim
later ones have a rim, but are brown
the changing traits overlap in time, but do not repeatedly appear and disappear
if the pots can be put in order like this, the order may represent change over time
but which is oldest and which is youngest?
you need some stratigraphic information, radiocarbon dates, etc. to get the direction right
a good seriation might be sensitive to changes over shorter periods of time than you could ever resolve with fuzzy radiocarbon dates...
Components
sites may be occupied at different times
each distinguishably different occupation is called a component (or, sometimes, an occupation!)
certain artifact types may be "diagnostic of" certain components
so these function as "index fossils", allowing us to link components of different sites together as being contemporary and, maybe, related by ethnicity or other traditions
the comparable components at many sites in a region may be lumped together as a "phase"; others might call this a period or a culture
these are often studied as if they were homogeneous units
but we can't forget that that is a simplification
in general, we consider different sites of the same phase to be contemporary
but in reality, we know that a phase is longer than an instant
so different sites of a single phase, while roughly belonging to the same time, might in fact have been occupied first one, then the other
Thomas talks about what the appropriate "unit of analysis" is at the site level, the regional level, etc.
but this actually depends on the question you are asking
it might be useful to compare different phases in a region to get a broad idea of change over time
but it might also be useful to compare different components within a phase in order to get an idea of the degree of specialization or variation within the phase
or it might be useful to look as different houses within a component to get an idea of family status differences or other patterns
Most of this has been about creating chronologies
chronologies are just the background for what we really care about: cultures and how they change
to create them, we intentionally ignore other interesting variation
this is what Thomas means on p. 117 when he says that "variability not attributable to time (and/or space) is irrelevant"
it is only irrelevant in that it does not help us with the specific task of developing a chronology
they are a necessary first step, but are not the story in themselves
who cares about the order and duration of pottery styles in southern Peru?
what we care about is how people lived, how societies changed, moved around, etc.
the sequences of pottery styles, point styles, etc. help us put things in order
they also may tell us something in themselves about what was going on
changes in points might indicate changes in subsistence practices
changes in ceramic decoration might reflect changing ethnicity, population movements, religious ideas, etc.
but that is a different, separate kind of analysis that requires more complicated data and thinking
ultimately, we will use a region's cultural chronology as a framework within which to reconstruct its culture history, which is what really interests us
actually, some would say culture history is still not the real point
who cares about how the political organization of southern Peruvians changed from 100 AD to 1400 AD?
some people might find that interesting or relevant to themselves; others might not
they would say that abstracting general processes of culture and culture change from the comparison of many culture histories is our real goal
they care about figuring out why cultures change; how leaders come to power; why complex states collapse, etc.
rather than any particular examples of those things, which are only clues to figuring out the overall pattern