Dear Folks
I totally agree with Jacqui- it really depends on your research
questions and the amount of time you have to look at the material- plus
the total NISP you have to work with. Plus the sort of deposit you are
working with and what you can reconstruct of its taphonomy (especially
rate of deposition- swift and "catastrophic" or gradual/episodic and
"accretional"). The house floor/activity area vrs midden scenario is a
common one- the floor deposits are probably mainly about behavioral
information (room use etc)during occupation or abandonment but the
middens tend to provide the big samples and the economic "big picture",
and it may make little sense to worry too much about horizontal
distribution of fragments in such a tertiary context (thrown away twice
at least). In the Norse sites in Greenland, it took me forever to
finally realize that the near absence of dog bones (but lots of canine
tooth marks on other bones) in the middens but the concentration of
sometimes cut and sometimes partly articulated dog bones in the last
floor layers meant something (sad) about final days behaviors vrs normal
patterns of consumption. In retrospect I would have been lots more
careful to treat indoor floor deposits as different from midden layers
right from the beginning.
Lumping gives you decent sample size (and somewhat amalgamates all the
sort term events which may in fact skew the overall patterns of the
deposit as a whole), but likewise this destroys your ability to see the
very short term activity-related stuff. Plus of course some short term
things skew the whole if averaged in- at the Svalbarš midden in NE
Iceland we had a single layer of extremely dense whale bone (all cut up,
probably result of a single event of outdoor whale bone working
following a nearby stranding)that held about 90% of all the whale bone
in the whole archaeofauna- just averaging this layer in with the rest
would certainly give an unrealistic impression of whale use at the site.
Likewise we have encountered concentrations of nearly whole brown trout
(with tiny gastropod gut contents in close association)
Which look like somebody's catch spoiling before it could be cleaned,
and the whole lot dumped on the midden....averaging this deposit in with
the rest of the layers around it would certainly boost the trout%! I
suppose this is another argument for close coordination between
excavator and analyst. We ask diggers to look out for articulations or
just lumps of bone and bag these together with an appropriate label so
we can have warning later in the lab of potential issues of this sort. I
think it is also always worth doing a check on context for anything that
looks unusual to see if those 395 unicorn bones all in fact come from
the same square and layer, or if they are pretty evenly distributed
across the site in time and space.
If your recording system codes for things like burning, fragment size,
animal chewing, cut marks and the like it is definitely worth doing some
numbers crunching and graphs to see if your contexts are actually
comparable. If area A is all calcined bones and a few mammal teeth with
most fragments smaller than 2 cm and area B is mainly 5cm and up,
unburnt bones with lots of fish and mollusca preserved one might think
twice before you average their bone counts together. Also if your field
people regularly record changes in pH (I have seen some impressive
shifts over short distances on some sites) this can really help in
flagging "best not lumped" contexts with radically different probable
taphonomic histories.
If people are interested, there are some excel templates on the "NABONE
8" package included in Sophia Perdikaris & Kate Krivogorskaya's FISHBONE
CD (more being made today- available free) made up for bovids, cervids,
pig, and seal with as much bone density/ MGUI etc. info on them as
seemed possible that may be of use in doing context-specific
investigations. At least you will rapidly find out if your context A is
all high value meat bones or not....
This is an interesting thread, many thanks for giving me a chance to
get my mind off our election.
Best
Tom
Thomas H McGovern
Professor,
Dept of Anthropology Hunter College CUNY
Archaeology Coordinator,
CUNY Doctoral Program in Anthropology
Coordinator, North Atlantic Biocultural Organization
Address:
Anthropology Dept.
Hunter College
695 Park Ave. NYC 10021 USA
tel. 212 772 5410 fax. 212 772 5423
[log in to unmask]
-----Original Message-----
From: Analysis of animal remains from archaeological sites
[mailto:[log in to unmask]] On Behalf Of Jacqui Mulville
Sent: Monday, November 08, 2004 4:56 AM
To: [log in to unmask]
Subject: Re: [ZOOARCH] Harris and zooarchaeology
Florencia,
I think the true answer to your enquiry lies in thinking about what you
are
trying to find out. For general conclusions about the changes in say
economic
strategies over time, the large sample sizes provided by grouping
together
individual contexts into chronological units are useful i.e. enough data
to say
something.
However if you want to get into the minutiate of your site it is
possible to
analyse each context to find out what bits of which animals are
deposited on
different house floors or indeed in different layers in pits. For this
you can
find out how much of each species is present using the MNE and MNI
(after Brain
for example). You also need to think about what exactly does each
context
represent, for example how long did it take to form and where did the
material
come from?
This sort of analysis can become very time consuming and involve rather
small
sample sizes. We are doing this type of analysis for the floors we find
with a
series of Bronze Age and Norse households, and for example we can find a
particular floor type assemblage, lots of fragments and the only
identificable
bits smaller dense bones such as phalanges. Middens on the other hand
contain,
as you would imagine, a much wider range of elements representing food
and
slaughter waste.
At context level I would be reluctant to try exprapolate the changes in
relative MNIs between floors to changes in herding strategies - rather
would see
it as differences in processing and deposition. The comparision
between say
floors and middens in general can also be interesting.
One could argue that, in a site consisting only of house floors, our
data
indicates that floors would not acurrately reflect overall economies, so
why try
to use this data to find out about them?
Well we have to a. work with the data we have and b. recognise the
limitations
of our data.
Regarding differences in taphonomy - I have neatly avoiding this by
working in
somewhat monotonous sites recently i.e. all one substrate. Again you
can become
involved in grouping pits/ditches/layer etc - but of course within each
of these
categories of deposition there can be a range of conditions - are all
pits equal
in pH, erosion or fragmentation? So at this point you can ask your
archaeologist for pertainent information and/or try quizzing your data
to look
for differences in preservation etc. In reality you need do think about
both.
Finally all of this is of course limited by time, which of course means
money.
Usefully you could examine some of the more recently published bone
reports to
see the approaches people take to urban data - e.g. York, Lincoln,
London etc.
Anyhow I hope my monday morning work avoidance helps. And feel free to
ask any
more questions. It a while since we had a quantification discussion.
Jacqui
Jacqui Mulville,
Honorary Meeting Secretary,
Prehistoric Society.
Lecturer in Bioarchaeoloy,
School of History and Archaeology
Cardiff University
Cardiff
CF10 3XU
Tel: + 44 (0) 29 2087 4247
Fax: + 44 (0) 29 2087 4929
|