Many pasts? The words we use
YOU! (Archaeology and the Internet)
see also:
(Descendants, archaeologists and others)
(names, genealogies, etc.)
The LJP Historical Society
(about community collaboration)
Links to other sites
Search (names, places, etc.)
Table of Contents
(lists all site links)

Levi Jordan

Interview with Robert Harris

Looking at the "Shell Carver's Cabin".

July 1998

Introduction, by Carol McDavid

I conducted this interview in the summer of 1998, in the Anthropology Department at the University of Houston. It includes the tables and drawings that we discussed then. These tables are also included separately – if you want to see them without the interpretations of them, go to:

Quantities of Artifacts in 5S/55W

Artifact Types in 5S/55W

Tools in "Shell and Bone Carver's Cabin" associated with carved shell and bone

Classification system used at Jordan Plantation (large file,159K- may take some time to download)

Page of unit notes taken during excavation

(sorry you have to scross sideways on this page; it's because there are some tables that take up lots of room)

Robert Harris

The gray-shaded unit (5S/55W) is the unit that Robert uses as an example for much of this interview. Here is a blow-up of the subunits in that unit

Subunit chart. Each 5' x 5' unit is subdivided into 25 1'x1' units as soon as the artifact density increases. (See Method for more details).

[after some introductory chit-chat]:

C: So, are you building on material from Doreen's [Doreen Cooper's] thesis of ten years ago, in terms of how she started looking at this whole abandonment question?

R: Yes, I've supported most of what she did and also found an important difference between what she was trying to do and what I am trying to do.

C: And what would that be?

R: During that initial period of time, Doreen and Ken's objective was primarily to determine if it [sudden abandonment] did indeed occur. That is, did people leave behind stuff somewhat quickly (things that you wouldn't normally leave behind) and in positions close to where they would have been left? She tested that by going through a number of five by five units and taking a random sample of subunits within those. She was able to show clearly that abandonment occurred – and did so with only two years worth of data available to her. I'm just trying to take it a bit farther, and to look at a large, contiguous area – a whole 25 by 25 foot area – to see what we can learn from looking at the precise, 3-dimensional boundaries of the deposit.

Where you can go from that, if it can be done, is to start to be able to look at the sub-floor deposits too. For, as much as we talk about the abandonment of this site, some of the best interpretations have come from the sub-floor deposits.[see curer's cabin for example].

[Among other things, Robert is attempting to "tease apart" the floor surface distribution from the sub-floor distribution. Ken has his doubts about whether this will be possible (with artifacts, anyway) though he does think it is an interesting question well worth examining. So stay tuned for the final results of this study!]

Anyway, we have these abandonment zones, and we have sub floor deposits beneath them. They can look a lot alike if all you've got is a chart of artifact counts to look at. How else can I tell the difference between the two? Well, I started looking at the variability of the artifacts, not just the artifact counts.

C: OK, so we're looking at a chart for the counts of the artifacts in 5 South, 55 West now.

The numbers indicated for "subunits", in the chart below, correspond to the locations on the site map above.

ARTIFACT COUNTS, UNIT # 5S/55w. See Method for a fuller description of how this site was excavated.

Level Subunit 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 in 5x5's Total
1                                                     32 32
2                                                     36 36
3   3 25 13 11 12 17 35   19 14 6 7 9 2 2 9 4 3     4 3         198
4   34 30 25 27 28 56 32   43 68 27 29 48 38 17 31 28 38 4 23 12 36 34 11 3   722
5   3 97 18 30 6 13 12   20 26 16 31 35 70 45 56 76 95 65 13 33   9 38 28   835
6     8 1 10 5 3     3 12       2 13 18 23 24 29 49 5 39 52 40 61   397
7     8   3           1                 1 5 1 3 23 24 6   75
8     2   1                 8   1 2       2       1 5   22
9                                                     17 17
10                                                     15 15
Total   40 170 57 82 51 89 79   85 121 49 67 100 112 78 116 131 160 99 92 55 81 118 114 103 100 2349

R: Yes, you can see here that levels three through eight were dug in one by ones. And then levels one, two, nine and ten were dug in five by five's. Now, in terms of the whole five by five unit, you can see pretty nicely it looks like there's an floor surface there, or some other reason why there is a high density deposit on some levels…

This table is just counts of the artifacts. After we look at this we'll look at variability. By the way, the cameo was found in…subunit 23, level six.

C: There was a lot of stuff in that subunit – Fifty-two artifacts?

R: Right. The highest density for that whole subunit was in level six. But look at level 5 – it's really a low count, 9. And subunit 22 has no artifacts at all in level 5. It looks to me as if someone may have dug too deep in level 4 – so, was it you that dug too deep?!?

C: Possibly, possibly!

R: When I see one subunit that has this big shadow [or really low artifact count] in it and all the other ones are behaving normally it makes me think that maybe there's something wrong with the depths...that somebody may have dug too deep or something. Now let's see. It's interesting that subunits 19 and 20 aren't near where any kind of wall should be but for some reason they started to increase in density at a deeper depth that the surrounding subunits.

C: From my notes it appears that Robert [my digging partner] must have been digging in that side of the unit, and we don't have his notes handy to see what he said about it. My notes do show that we were going to close the unit at level eight...and reopen it the next year…but then it looks like we went ahead and excavated down to level 10…what happens in the chart after level eight?

[This is a section of the notes Carol took when excavating this unit. For a 63K picture of some other notes from his unit, go to NOTES.

R: Not a whole lot. It just went down two more levels, in 5 x 5's.

C: If we had stopped at level 8 like we originally planned, instead of going on…and then it was re-opened the next year….do you think it would important to take account of that during your analysis? I mean, what if a bunch of cows tromped on it in the intervening year? Even though we covered it up well with plastic sheeting and so forth?

R: Yes, and I'll tell you why. Sometimes bizarre intrusions happen between when a unit is closed and then completed later on…after being re-opened. What happened some of the time is that people stopped, say, at level seven. They dug down as deeply as they needed to go. And they measured properly and all. There is this phenomenon that I like to call – well, I don't know if it is a real word, but I'll call it isostatic rebound – when there is no longer a ton of weight on some piece of ground, it will tend to rise back up over a period of time. For example, most of the interior of Greenland is below sea level because it's got a big, heavy glacier on top of it. If the glacier went away, it would rise back up some because it wouldn't have tons and tons and tons of weight on it pushing it down – So, if you've got a big square hole in the ground and you leave it for a year or so, it would tend to rise back up, a little bit anyway. So, what happened was that, when going back into the unit, they re-measured and said... "Oh my, they didn't finish level eight."

C: I remember that did happen sometimes.

R: And the majority of the time what happens is that the new excavators will go through 25 more subunits of level eight, taking it down that tenth or so that it had sort of popped up – which can be very confusing to match up in the lab.

C: OK, back to our charts – "Quantities" and "Types." So – to look at variabilities, you go to the "Sum of Type" information?

R: Right, what we are looking at now is how many types of artifacts there were in each subunit, by level. This one here [Subunit 10] is a good example. It goes from three types on level three, and 16 types of artifacts in level 4. Sixteen different types of artifacts in one square foot of soil, only 1/10 of a foot deep!

Artifact types in 5S/55W (see subunit chart above for spatial arrangement of artifacts)

Level Subunit 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 in 5x5's
1                                                     10
2                                                     10
3   3 7 6 3 7 10 5   6 3 4 4 4 1 2 3 2 1     3 2        
4   5 6 6 6 10 9 8   10 16 10 6 13 10 4 8 6 10 4 7 3 11 7 5 2  
5   2 19 7 6 5 4 3   5 6 7 10 10 12 10 9 13 12 9 5 10   6 7 8  
6     6 1 1 3 3     2 6       1 6 7 7 5 6 12 3 8 12 11 9  
7     1   2           1                 1 5 1 1 3 4 3  
8     2   1                 1   1 1       2       1 3  
9                                                     4
10                                                     1

C: And subunit 20 has 12 types of artifacts.

R: Right - what I did is I made all these different charts for every single unit in the area we had already identified as being within the cabin walls. I made charts of count, of variability, and also made charts of weights of artifacts and a couple of other things. I looked at weights because one hypothesis is that the SMALL artifacts could have passed between the cracks in the floor boards. Large ones would, one would suppose, sit on top of the floor. Unless they were purposely buried...[as they appear to have been in the curer's cabin].

C: Do you have buried stuff in this cabin?

R: I think so. I want to go back and redo all my abandonment checking because I've got a little more data now that I've measured the weights of artifacts individually, rather than as averages. Also, looking at artifact types can be tricky, because I have an idea that certain types of artifacts should only be present within an abandonment zone. Looking at several different types of data independently keeps me out of the interpretations, as much as that is possible, at least.

C: By "keeping you out of the interpretations", do you mean that you sometimes have to consciously screen out certain types of information that might influence you, as a cross check? Do you think you are kind of testing yourself all the while as well?

R: Yes, every single time I say, "Here's where the abandonment stops" I have to do that. Sometimes, it's a pretty straightforward decision. Most of the time it's not.

C: On a related question – how important is note-taking? Do you find yourself going back to the notes done by the original excavators??

R: Yes…when the notes are detailed, it really helps. For example, like I described earlier - sometimes I'll have what appears to be a "spike" in an artifact count for one level, then it's really low in artifacts or even sterile on the next level, and then a spike again... and then I start wondering – do I have a really big subfloor deposit? – or, did an overzealous excavator dig out two tenths of a foot for a level instead of one?

C: So when someone goes too deep, it can screw things up??

R: Yes! That's when good notes really count – when the crew chief will note that somebody dug too deep, that will sometimes explain things. You know, when you're in the field, one of the problems with field archaeology is that you're DIGGING. You're not thinking about these kinds of problems, usually, unless you've done a lot of archaeology. And most of our excavators are volunteers or students, and they aren't used to measuring levels and so on. They sometimes dig too deep. I certainly didn't start thinking about these problems until I had to start dealing with data.

C: Do you think 5/55 is fairly typical of this cabin area?

R: It's one of the more straightforward ones. It's typical of what a straightforward abandonment profile looks like, or at least a floor surface profile. They aren't necessarily the same thing...what I am trying to do is to isolate what was most likely left by the occupants versus what was most likely thrown in afterward, as trash or whatever, if that did in fact happen. Once I've got the floor level established I can start to look at that.

Let's look at the Sum of Quantities table again.

Level Subunit 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 in 5x5's Total
1                                                     32 32
2                                                     36 36
3   3 25 13 11 12 17 35   19 14 6 7 9 2 2 9 4 3     4 3         198
4   34 30 25 27 28 56 32   43 68 27 29 48 38 17 31 28 38 4 23 12 36 34 11 3   722
5   3 97 18 30 6 13 12   20 26 16 31 35 70 45 56 76 95 65 13 33   9 38 28   835
6     8 1 10 5 3     3 12       2 13 18 23 24 29 49 5 39 52 40 61   397
7     8   3           1                 1 5 1 3 23 24 6   75
8     2   1                 8   1 2       2       1 5   22
9                                                     17 17
10                                                     15 15
Total   40 170 57 82 51 89 79   85 121 49 67 100 112 78 116 131 160 99 92 55 81 118 114 103 100 2349

R: We're looking at subunit two, levels three through five. And then the total artifact count is 25 and 30, then 97 and then it drops to eight. Between level five at 97 artifacts and level six at eight artifacts, this huge drop tells me that something is going on. There could be just a bunch of trash thrown in there. Any floor surface can accumulate stuff. So at this point I know I have found the floor surface. I have not found what created it and why those artifacts are there. So then I have to look at variabilities.

C: Wouldn't you argue for more variety if it was a random distribution – a trash heap?

R: Across a broad area, sure, and that's where the third test for abandonment comes in. I know that there is probably a floor surface there between 97 artifacts at level five and eight artifacts at level 6. All this vertical information from counts and types is telling me is where the floor surface is – and I think that's all that the vertical information is really good for, in the final analysis. What I need to do next is to compile all of the proveniences together that make up that floor surface horizontally and then apply the tests specifically designed to identify an abandoned surface on that floor surface.

C: So have you done that?

R: I'm starting to. I need to look at the spatial distribution of the actual types that are represented there on the floor surface. It's a horizontal analysis. I think the test for the floor surface should be solely vertical and then I think the test for abandonment should be a separate entity upon whatever you've established as the floor. The abandoned remains are supposed to represent where these items were when they were left behind by the former occupants-- and in the way people really USE households, and floor space, they don't have curatable objects spread evenly across the floor space. [You have open spaces, storage spaces, and so on.] What happens if you go through and you look at one subunit at a time to catch the vertical variation in the floor height. If you are looking for curatable artifacts in every subunit, they are not going to be there and it can give you a false negative on the abandonment test.

C: Yeah, because you wouldn't have curatable artifacts everywhere on your floor.

R: Right. A floor surface, by definition, must span the entire area of a cabin. Therefore, either the entire floor surface was indeed abandoned, or the entire floor surface was not. You cannot say that only 40% of the floor surface was abandoned. That would be silly, like saying that you abandoned your house, except for the west half of the kitchen.

C: You mean you can't have it both ways?

R: Right. But if you are looking for whole, usable, curatable objects as a test for abandonment in each and every 1 foot by 1 foot subunit, that is exactly the kind of result you will get. Not every single subunit is going to have that kind of stuff, even though the area that it represents was abandoned just as much as the one that the unbroken eyeglasses were found in.

C: Unless you have it abandoned and then used it for secondary refuse?

R: True. But that still shouldn't change where the curatable stuff is. If trash was thrown in on top of it, it should be randomly distributed, and in any event probably did not knock the stuff that was in there already around too much.

I'm not as worried about the possibility of intrusion as I once was – even if some stuff was scattered in on top of the abandonment items, I'm still reasonably sure that most of the abandonment items are still pretty much in their original places.

C: Is this because you have hand-mapped all the artifacts? Don't you just use the photographs and drawings we did in the field?

R; Sure, but we don't have as many as I need, and sometimes the maps I make now are clearer. Since I still have the information from the 1x1 excavation method, I don't have to just rely on the drawings. Have I explained what I am trying to do with the vertical information?


C: Okay, the abandonment tests are count, variability…and?

R: And the presence of curatable items, things you wouldn't find in a trash can, and the size of artifacts (as approximated by their individual weights), as well as their orientation in the ground. What led me to this was the problems I encountered in my preliminary analysis for the conference [referring to his SHA paper]

For that paper I was testing for abandonment and floor surface at the same time. I had charts for counts, charts for variety, charts for weight, and charts for personal and curatable items. The problem played out something like this - I would have a subunit here that met only one of the criteria for abandonment. I've got subunits all around it that meet two, three, sometimes all four criteria. How do I decide how many criteria are needed to declare an abandonment? And that's when I realized that I've got to look for the floor first. And THEN test abandonment. Once I find the floor, it doesn't matter if I've only got five curatable artifacts. If it meets all the other abandonment tests, then the whole floor was abandoned.

C: But you've also got to account for all this other stuff that's not curatable, that's more randomly distributed – so you tease the two tests apart and look at it independently?

R: I think that if I can do that it will give me a much cleaner end product. I think you have to find the floor surface first, and then test for the formation process that created it, for what's really there, horizontally. What I do is I establish the floor surface and then I check out the spatial patterns and then I check for curatable artifacts --

C: ... and distribution...?

R: ... and associations, things like that.

C: So after testing things vertically, you started to look at them horizontally...and you tried to use [the geographical graphics computer program] Surfer to map things (as we discussed during a break in the tape), but you found it really wasn't suitable for your type of study. Mainly because it tended to map things along contour lines, and didn't account well for areas in which there were NO artifacts...and it tended to put in big spikes of artifacts where there were only two or three artifacts. So then you went back to a hand-mapping system, and mapped each artifact type in by hand, over the entire cabin area. Right?

R: More or less. I combined several of the basic classification categories into larger ones that I thought would have some utility for looking at overall horizontal distribution. I also wanted to confirm the location of the wall between the cabins.I wasn't sure where it was, since the wall wasn't made out of anything that left traces of its construction. [Note: there was an hypothesized wall location, based on the dimensions found in cabins in other parts of the site, plus an artifact "shadow" in this the same area on the drawings made during excavation].

[Go to categories if you want to see all of the categories used for the Jordan Plantation archaeology. Be sure you have time for it to download, because it is a 150 K image.]

C: Do you know where the door is here? Do you have any of the locks from this cabin?

R: I've got some locks. I'm starting to try and look at where the doors were. And I think there are windows over on the south side…what I then noticed was that all of the stuff that I thought was associated with shell carving activity was on the north side of the cabin.

Here's my hand-drawn diagram of shell and bone artifacts that I can verify. [we will try to include a copy of the hand-drawn map in the future] All of these things occur on the southern side of where Ken and I think the cabin boundary is. These are the bone and shell artifacts that I could positively identify. And other things that could possibly have been associated with the shell carving industry.

This is a map of the objects associated with shell and bone carving that are located in the same area as carved shell and bone objects themselves. The line inside the cabin area is an hypothesized interior wall, which Robert is also testing for. The fireplace is shown on the right side of the drawing.

C: And these things don't show up in the "living area" where the fireplace is – with, perhaps, one or two exceptions? For example – do you have pocket knives elsewhere?

R: Yes…and some other tools do occur in what I think is the the living area. But although I have the same varieties of tools more or less on both sides of the cabin – for example, there are some high chisel and file concentrations in one corner – the types of tools are more or less the same on both sides. The key part is what they are associated with.

C: You mean, in this northern part of the cabin, they are close to all the bone? There's no carved bone in that south part of the cabin?

R: None that I could identify. There is an area where I saw a second concentration of tools on the southwest side and it's got some sharp devices – some files and the shell blank was found there. This may have been an area where blanks were formed – where they were prepared prior to the actual carving process. Jorge has evidence for two-stage manufacturing process in the munitions area, and that could have been happening here too.

C: What about the southeast corner with the chisels?

R: I'm not sure. I haven't looked at the full association of those. We could also easily confuse where stuff was stored versus where it was used, and I also have to look at what types of refuse was there too.

What I am doing now to further test where that northern wall is – I've got 91-some-odd types of ceramics. I've got at least 60 types of glass. What I've been doing is looking at where these types cluster – and so far the vast majority of these things occur only on one side or the other of the hypothesized interior partition line, other than one or two miscellaneous pieces.What I am going to do now is to look at each individual distribution of some of the more sensitive artifact types, and then draw a line on these maps where I thought the [interior] wall might have been, according to each map.

C: This kind of analysis takes quite a bit of time, doesn't it?

R: It's hard to do but it's worth doing. I'm getting pretty good interpretations from it.

C: How would you want to improve it?

R: I'd want better drawings, profiles and note-takings. I wouldn't want any smaller proveniences.

C: No smaller than one foot by one foot?

R: Sometimes a one by one is too small. I sometimes have to combine it into something larger to get a meaningful image - but the advantage of doing it is because when I DO need the information it provides I have it – if all I had was big units I couldn't break them up into smaller ones.

C: So you wouldn't be able to even KNOW there was an activity area without it – and Ken wouldn't have found the curer's kit or anything else in that cabin in such detail, and so on? Which is why it might be hard to do this kind of analysis with the seamstress's cabin? [which was excavated early in the excavation, and was mostly excavated in 5 x 5 units.]

OK, you are going back to the artifact distribution and testing for the wall right now – you're not just assuming that the wall is actually there….

R: Well, I have been confident enough to provisionally say it was, but I want it much more thoroughly documented for my final thesis.

C: And that's what you are doing with the ceramics and the glass?

R: Among other things.

C: So you've got some other things in mind as well?

R: I've got some glass types that may help to date the deposit. I'm also looking at how was it scattered, whether it was up against the walls...and, as I mentioned earlier, I'm looking at the weights of individual pieces to help me figure out the floor levels.

Ginny McNeill Raska
Hazel Austin
Dorothy Cotton
Kenneth L. Brown,Ph.D.
Morris Richardson
Julia Mack
Cassie Johnson
Sarah Martin
Carol McDavid
Donna Gregurek
Marci Naquin
Mary Barnes
Mary Lynne Hill
Jorge Garcia-Herreros
Robert Harris
David Bruner
April Hayes
Doreen C. Cooper
Rebecca Barrerra
Cheryl Wright
Kris Brown

We need your help to make this site better! Please tell us what you think by going to:
Questionnaire #1 (Adults) * Questionnaire #2 (Kids)*
Feedback Form*

HOME * Conversations * Words * Archaeology * History * Ethnography * Community * Media * Descendants & others * People in the Past * Kids * Bibliography * Definitions * Links * Maps * Search * Table of Contents
For information about this site or this project, contact or .
Carol McDavid 1998