Typologies in pieces: Fragmentation, reconnection, and granularity

 The summer of 2012 was crucial in my thesis – I had finished my theoretical framework and a couple of chapters. I was about to tackle the most grounded part of my research: wading through a lake of approximately 40,000 bits of broken pottery.

 The purpose was to make sense of that material to understand better how people in an Eastern Mediterranean island 3500 years ago decorated their stuff. But how to make sense of 40,000 bits of pottery? Typology comes to the rescue. Over some glasses of potent raki, my supervisor shared his wisdom: “Well, you know. There are two kinds of archaeologists: lumpers and splitters, and you have to decide which one you are.” I sat with this. Is it better to be a lumper or a splitter? Put things together or take them apart? As a matter of caution, I thought, under conditions of uncertainty, it is far better to be a splitter: after all, you can recombine something that has been split, but once lumped, splitting is nearly impossible. 

Granularity and typology

This moment came to my mind the other day as I reflected on the concept of granularity. Optimising granularity, usually towards a finer grain, has been a heuristic of managing complexity for a long time. I found references to it on the blog dating back to 2007, and less than a month ago, Dave Snowden was reflecting on the concept through the lens of different Welsh words for “granular”. Archaeology also brought to the forefront granularity’s connection with typology. By typology, I mean a sense-making scheme that helps us look at things from different perspectives, not a device for permanent classification. So, in terms of granularity, the typology we use both has granularity and enables or encourages us to engage with finer or coarser items. It is, then, helpful to look at both ideas in parallel. 

Some heuristics

My musings on granularity and typology led me to a series of heuristics I learned from dealing with those 40,000 potsherds that can apply more broadly. 

1. If in doubt, go lower.

As my initial instincts told me when confronted with the question, more finely grained means safer, especially once the original material becomes too much to handle except in aggregate. After my pots leave my tables, they return to a museum basement and become numbers on a spreadsheet. If I decide that decorative lines and decorative dots are the same thing, I can add them up and no harm done. But if I had decided in advance that they must be the same thing, and then I realised they’re very different, how would I separate them without access to the original material? The fewer things we know in advance, the fewer assumptions we can safely make, and the more granular the objects we’re dealing with should be. 

2. It is possible to go too low.

The lowest coherent level of granularity isn’t necessarily the lowest possible level of granularity. The key word repeated since the early days of the term’s use is coherent – and coherence is very much context-dependent. In other words, it takes an expert in the context to tell us what is coherent and what is not. In my case, the expert I would ideally need is the actual user of the pottery. Still, in their absence, we make do with those who have spent hours and hours immersed in it. The casual viewer might think that a slightly thicker or more vertical rim is a meaningless level of detail, while to the expert, that holds a lot of information. On the other hand, if the expert falls in love with their typology, we might end up with countless subcategories of rims and bottoms that mean little beyond our desire to classify the world neatly. To refer to an example I shared when teaching recently, if we are trying to manage diabetes, then “the body” might be too coarse-grained (which aspect of the body?), but going down to individual DNA bases is probably not particularly helpful either.  

3. Incorporate variety and diversity into typologies.

It’s helpful if your typology has several dimensions, or you can look across several typologies. We established that a typology is a way of looking at things from different perspectives. In that case, using more than one typology can become a way to connect perspectives and kick off the process of recombination – bringing together granular items in new ways. When trying to make sense of someone’s millennia-old household, I might want to consider shape; am I dealing with a jar or a wine cup? A plate for serving or a jug for pouring? However, the quality of the material could be another dimension: is it rough and tough and coarse, or silken-smooth? And for what purpose? Now, I can put those two dimensions together. I can start noticing exciting and novel things, such as an incredibly elaborate shape created in the crappiest material imaginable – information that no one dimension can ever capture because it’s about how things come together and giving ourselves the tools to notice that.

4. Connection through abstraction.

Concepts at a higher level of abstraction can help us draw diverse and valuable connections across typologies and items. The heuristic above starts hinting at that – how slightly more abstract ideas, like “shape” or “decoration”, but also like “celebration”, “ritual”, or “household” can bring together our granular material so that it goes beyond being “this pot” and “that pot”, but instead becomes a collection of items that work together, are active together to produce effects through their interaction: from all that comes together to prepare a meal (storage, preparation, cooking, serving) to holding a fancy party – and the purposes that party can serve. We looked at the lower coherent level of granularity, and abstraction now allows us to bring relationships back in the game, looking at how things together change one another and what the whole assemblage can do.  To connect back to some of our own work, this is exactly why SenseMaker allows people to relate their experiences to a network of concepts at a higher level of abstraction; new connections.

5. Hold your typology lightly.

Have you ever opened your cupboards and pondered the typological analysis of your mugs and plates? You probably haven’t because you are their actual user. You instead have different categories that are more meaningful to you, like “the good china” and “the mug that is the perfect size for Monday’s coffee”. But when your mugs end up on an archaeologist’s table a millennium from now, this archaeologist will have to use their own typology to make sense of it. And they must remember that this typology, like all our schemas, breakdowns, and frameworks, is not “the thing”. It’s a way of looking at the thing. We hold them not because they’re true but because they’re helpful. And we should always be on our guard against mistaking the abstraction for the thing itself.

.

Image credits

Banner image by Karol m on Flickr. This file is licensed under the Creative Commons Attribution 2.0 Generic license.

In-text image 1 from G. A. Reisner’s A history of the Giza necropolis. Vol. 2: The tomb of Hetep-Heres, the mother of Cheops, 1955, fig. 72. This work is in the public domain in its country of origin and other countries and areas where the copyright term is the author’s life plus 70 years or fewer.

In-text image 2 by Gary Todd on Flickr. This file is made available under the Creative Commons CC0 1.0 Universal Public Domain Dedication.