Before leaving for the field in May I arranged to meet with top-notch classical bioarchaeologist Sherry Fox. One of my advisors had put me in touch with her via email, insisting that I reach out to her as her research interests were so pertinent to my own dissertation foci, and she kindly agreed to meet with me for coffee. I had a plethora of ulterior motives for the rendez-vous besides my persistent caffeine addiction; I wanted to pick her brain about strategies for dealing with commingled and fragmentary human remains, get some advice on best practices for MNI calculations in prehistoric necropolises, and, as it turns out, to catch up on what life in Athens is like in the 10+ years it’s been since I’ve spent time in the city. One of the pieces of advice she offered that stuck with me was to take photos of all of my teeth.
At the time I remember thinking it seemed a daunting task to re-photograph the ~200 loose teeth I’d examined the previous summer. I’d be forced to make decisions about orientation, keep track of each individual tooth relative to its unique identifier, and manage to tag or organize the photographs in such a way that I’d be able to pinpoint a specific tooth without much effort if I needed to double-check an identification or re-assess pathology or age estimates. Should I photograph each individual tooth? Should I only photograph teeth with caries or pathology? Should I only photograph one view of the teeth (e.g. only the occlusal surface)? All in all, a lot of tedious, time-consuming work seemed necessary to come up with a useful system. Even thinking through the process was a little overwhelming. However, once I realized that one of my necropolises contained so much fragmentary human bone that a dental analysis was the only efficient way to garner information about age and MNI, I was forced to come up with a system, and quickly.
I’m sharing that system below, both because I found these sorts of guidelines and templates extremely useful when organizing my own initial research in the summer of 2013, and because I’m eager for feedback from fellow bioarchaeologists on how they handle these kinds of logistical quagmires.
The first step in my dental analysis entailed sorting every bag of bone for loose human teeth, or any fragments of alveolar bone that contained alveolar sockets (open or resorbing) or articulated dentition. While occasionally the site archaeologists threw me a bone (see what I did there?) by bagging all of the teeth in small, interior bags like in the photo above, for the most part I had to go through things by hand:
After I had all of the teeth pulled, my strategy was to:
1. Sort the teeth into groups by category and arcade: I began with the lower central incisors, and then moved through all permanent mandibular dentition, moving in a mesial –> distal direction, ending on with permanent lower third molars. I followed this up with all permanent maxillary dentition, saving developing permanent dentition and deciduous or developing deciduous dentition for last. To sum up, I kept the following groups separate:
- permanent mandibular dentition,
- permanent maxillary dentition,
- developing permanent dentition,
- deciduous dentition and
- developing deciduous dentition,
Then, within each group I organized the teeth in rows based on category (e.g. incisors, canines, premolars and molars), moving from mesial to distal within each group, so that the groups always started with incisors and ended with molars. This system meant that it was easy for me to compare all of the teeth in a given category to see if there were any left-right pairs that mirrored each other, since all of the teeth in a given group were from the same arcade and at the same level of development.
2. Arrange the sorted teeth into rows of ten on a tray: Private lab space is rare, and the chances of other individuals frequenting the desks and tables where you conduct your analyses are high. Accordingly, in the event that I needed to move teeth around, whether to facilitate the access of cleaning staff, clear space for behind-the-scenes museum visits, or simply reorganize my own work, I always laid out loose teeth on trays. My upper limit tended to be about 50-60 teeth a tray; any more than that became difficult to handle if the tray was accidentally jostled and re-sorting was required.
I stuck to rows of ten because I would input that many teeth into my spreadsheets as a group. For example, if I had a group of ten incisors I would generally move column by column, weighing them all sequentially, or examining them for calculus or caries sequentially, as I found this moved analysis along faster than inputting all fields for a single tooth. Analyzing teeth as a batch was particularly useful when they didn’t demonstrate pathology or were all at the same level of root development, as I could then drag and paste common cells like “Roots – apex complete”, filling in columns more quickly. However, I found that over ten teeth per batch became too confusing, and too difficult to backtrack through when I was unravelling a mistake.
Finally, the organization of the rows of teeth reflects their unique alphanumeric identifiers, which are sequential. I always started with the lowest number in the upper left-hand corner, with numbers increasing as the teeth moved right and down. Basically, the sequence would proceed along the same lines as the directionality of the English script, proceeding left to right and from the top to bottom of the sandbox. In the photo below, for example, the first tooth in the upper left-hand corner is T.01, the tooth to its right is T.02, and the final tooth in the sequence, the molar on the bottom right just above the photo-scale, is T.53. This method makes it easy for me to peruse photos and pinpoint a specific tooth – for example, I can tell that the large upper molar with the massive caries in the lower left-hand corner of the frame would be T.041 in this provenience sequence (I just checked my spreadsheet, and that tooth is indeed T.1.006.41, so thus far the system works, n = 1).
3. Prepare the loose teeth for photographs by using a sandbox: This was the part of dental analysis that required the most mental fortitude. Handling many loose human teeth is, frankly, an enormous pain in the gluteal muscle group. However, if you conduct your research abroad and access to collections is a one-time boon, such photographs may provide your only means for sorting out analytical quandaries or inconsistencies in your dataset that crop up as analysis proceeds. While I wish I was a rigorous bioarchaeological automaton capable of collecting thousands of rows of data with nary a single error, I know that mistakes have a Hydra-like tendency to spring up and multiply as a field season progresses, and redundancy (e.g. having your data “present” as both quantitative spreadsheet data and photographs) is one of the key safeguards against basic human error derailing the rigor of your research.
Despite my commitment to analytical rigor, I would rather chew off my own foot than photograph 3800 teeth individually, so I attempted to maximize both efficiency and information by photographing teeth in batches, from multiple views. I decided to take photos of the mesial, distal, buccal, lingual and occlusal sides of all teeth, using a sandbox to hold the teeth in position. This strategy was not the result of blind adherence to arcane principles of ‘thoroughness’, but rather designed to make the process of taking photographs require as little thought as possible. Orienting the teeth to take one photo so that, for example, all caries were visible, or all hypoplasias were apparent, would have required a lot more thought and consultation of spread-sheets, and when dealing with 175 loose teeth in a bag, I didn’t want to get lost in the minutiae of orienting and attending to each individual tooth. This strategy meant that I had an automatic system in place for photographing the teeth that (a) required very little thought to set up, and (b) would give me at least one view of every side of each tooth, thereby providing documentation of any caries, enamel pearls, hypoplasias, pathologies, etc., and (c) one view of the occlusal surface of each tooth, allowing me to go back and double-check attrition scores if necessary.
4. Photograph the teeth: I also needed to develop a system that I could easily replicate each time I took a photo, one that would also allow me to automatically identify what the orientation of the photograph was without relabelling it. I decided that the least painful strategy to pursue was to position each group of ten teeth in the sandbox as soon as they were analyzed, orienting them so that their mesial side was visible (as in the photo DSCN0128 above). I began with the mesial side of the tooth as mesial and distal are the trickiest orientations, that require knowledge of what side the tooth is from in order to position the element. This orientation was most easily achieved right after analysing a row of ten teeth, as it was simple to go back through my identifications and check siding – distal orientation then simply required flipping the tooth 180˚. After that, buccal, lingual and occlusal views were always easy to identify, even without remembering the side the tooth is from. Whenever a tooth was unsided, it was oriented as if it was from the right side of the arcade, for the sake of consistency. The maximum amount of teeth I would photograph in one session was about 60, as any more than that would have required sacrificing the level of photographic detail possible for each tooth. All told, this system took between 10-15 minutes for larger-batches of teeth, and the process got faster over time.
5. Organizing the photos: I kept a concise log of what I did each day in a word document on my desktop, and every day I input the order in which I took photos. At the end of every workday, I input all photos from that day and arranged them by provenience, using different designations for teeth, mandibles and maxillae. One key step to undertake before starting a project of this scope is to make sure your camera date/time is correctly set, because this can help you to back-track and figure out the subject of specific photographs. Combing through several days worth of photos is an arduous process, however, and makes it easy to misidentify photos, so I would strongly recommend loading and organizing photos at the end of each day of analysis, when you are most likely to remember what the photographs were taken of if your notes lead you astray.
As an example of how I organized photographs themselves, the teeth in the photos above were from bag 1/006, and were given unique identifiers using a sequential numeric system – hence T.1.006.01 – T.1.006.53. In my photo files, all three of the files shown below would be housed in a larger file titled “1/006”.
While this made for some frustrating afternoons attempting to flip tiny little deciduous incisors onto their distal sides, it also means that I can look at a sequence of photos like the one shown below and know that they are sequentially ordered according to that organization. Gaps in numbering systems occurred whenever I took a bad photo and deleted it, but the ascending sequence would always follow the sequence of orientations outline above. Accordingly, if I wanted to examine the lingual view of a specific tooth from 1/006, all I have to do is open the folder and go to DSCN0132, the fourth photograph in the sequence.
And there you have it – my tedious, but thorough, system for photographing thousands of loose teeth. If you’re a bioarchaeologist, osteologist or paleoanthropologist reading this blog, how do you organize and photograph loose dentition? What are strategies that you use for keeping dentition organized without going insane?
Image Credits: All photos were taken at the Museo de Jaén in summer 2014.