| Entry: | Working with Maurizio and the UC Merced team has been an interesting experience over the last couple of weeks and I have been remiss in my diary entries partly by accident and partly by design, we have been shifting a lot of room-fill from Building 89 and there has not been much stratigraphic incentive for comment, but furthermore I have been formulating my thoughts on what we are trying to achieve this season, so that I could set them down in this entry.
So I think the time has finally come to set down some of my observations on how it is all going. I think first however it might be a good idea to summarise what our actual goals are for the season. Starting with the ‘analogue archaeology’:
From my point of view the mission statement is simple: continue to excavate in the south sequence below the burnt [Mellaart] Level 6 Building 76 (mostly excavated by Mike House in 2009) and expose and empty the underlying (Level 7?) Building 89. From this perspective things have been going well. We took the decision to quadrant the structure, in an attempt to exercise a further degree of rigour to our stratigraphic control of the room fill - This has been problematic from a logistical point of view for a number of reasons, mostly related to the implementation of the digital methods (see below), although stratigraphically it inevitably causes some problems when things start to get complex. For example the presence of an ‘annex’ in the northeast corner of the building [Space 378?] has locked the northeast quadrant and has forced us to invert them whilst keeping a running section. It is unclear how the implementation of the quadrants will work down to the floor, but I am happy that we will get a complete cross section of the room by keeping the running section...
From the point of view of the UC Merced’s ‘digital archaeology’: As I understand it they are aiming to simultaneously work alongside our (Catalhoyuk’s) recording system by testing a number of different photo-modelling and laser-scanning techniques. Apologies to the team if I paraphrase this badly, but I think their aim is to test the effectiveness and accuracy of some these technologies whilst at the same time trying to implement an efficient and effective workflow for the capture of a 3D digital archive. To that extent I am very excited about the whole project and it is interesting to me to see how they are going about this. A number of things have come to mind over recent days which I think need to be discussed (in fact we do discuss them a lot on site, but in the interest of being reflexive I thought they might be aired in the diary as well)
At first I was a little skeptical as to whether this approach was going to be workable - I was concerned that the digital model was going to reflect the archaeological process. Would it be too rigid? Too slow? How would we edit and alter the archive on the fly as interpretation in the field changes whilst we dig things? Indeed I will come back to the latter point as I think this highlights one of the key issues in this whole process.
Nevertheless after several days working with the UC Merced team, I am struck by how the implementation of a 3D digital workflow, rather than working against the goals of our single context recording methodology, actually complements it very closely. The process is fairly flexible, and certainly not as slow as I expected (it takes maybe 20 minutes longer to record than our analogue approach - longer perhaps if the scan is complex and needs more detail - but certainly not the hours and hours that I was expecting).
Critically it is clear to me that what the ‘digital archs.’ are after is a record of the stratigraphic interfaces, be they positive or negative. Ultimately this is exactly the same as what we are trying to achieve our the more ‘traditional’ recording. I think perhaps our perspectives differ, I feel the ‘digital archs.’ are primarily after complete spatial control for their models, which forces stratigraphic rigour. However I think my priority is complete stratigraphic control which is best implemented by careful spatial recording in single context (aka. full spatial control), however this is not the only tool we have to achieve stratigraphic control - sections being the obvious alternative. Here is where I come back to quadrants, since it is clear that whilst the quadrants might be logistically harder to implement than excavating completely in plan, in terms of stratigraphic control I recognise that they do act as a mechanism for checking the subtleties of the excavation process. However whilst the ‘analog archs.’ can work around this, I can see that it clearly throws up issues for the ‘digital archs.’ Leaving in sections means that they do not get to see their interfaces in their entirety and this make life harder in terms of stitching point-clouds together and making deposits whole in the virtual 3D environment.
This brings me back to my earlier question of interpretation ‘on the fly’ because as Nicolo pointed out today the digital recording methods are passive in the sense that they strip out the interpretive process that drawing a stratigraphic unit (and editing that unit when you change your mind) adds to the analogue recording methods. To my mind it throws up a number of questions which as a team we need to address. The first and most important is what precisely are we trying to achieve, or record with these technologies? In a sense I am being deliberately provocative here as I think I already know Maurizio’s response to this (but that is for him to outline I think). However in my view it comes down to two options, which by no means do I see as being mutually exclusive: Are we trying to generate a 3D model of the stratigraphy as we understand it, stratum by stratum, or are we simply recording the excavation process as it unfolds in 3D? These are subtly different concepts but I suspect that the implication of each is very different for the workflow of these technologies.
The former is very hard to implement, if we are going for an accurate representation of the stratigraphy (as we in the field understand it) then ideally we must define and dig it correctly first time, we must understand the interfaces before we record (scan) the deposit or cut so that the model is correct. As we know this just doesn’t really happen: we change our mind, we edit plans in the field (and in the post-ex), we get things wrong and have to reconstruct the stratigraphy once we realise the error, and sometime we implement an excavation methodology which deliberately compromises our spatial control in order to exercise stratigraphic or sampling control (sections, quadrants, etc...) all of which risk compromising the digital spatial model or at least adding hours to the post-processing (which can take a while - based upon my observation in the evenings).
If however we don’t set out for the perfect model, but rather aim to make a record of the excavation process as we go then I think we have a vastly more straightforward workflow and this (based upon my observations) appears to be the way we are going with this - effectively modelling each unit as we come to a point where we would ordinarily plan it. I wonder if doing this with care and consideration would eventually lead to a working model of the stratigraphy anyway, since ultimately this is what we are trying to achieve by excavation anyway surely? Thinking of it in terms of recording the excavation process, rather than modelling the stratigraphy might lead to some redundancy and difficulties in the 3D data capture (arbitrary layers, recording and acknowledgement of interpretive errors in the field, recording of sectioned layers, rescanning areas when mistakes have been made), but I wonder if this is a more realistic way of thinking about this digital recording process?
Ultimately this leads me to question the resolution as well - we talk a lot about microns which seems weird to me as I tend to think about archaeology on metres and centimetres (the latter being - let’s face it - the difference between a blunt pencil at scale 1:20 or a good heavy cleaning). Since we don’t dig in microns my question is it worth recording in microns - perhaps for ‘harder archaeology’ such as architecture - but is it necessary for the ‘softer stuff’ (the dump with very diffuse boundaries)? The reason I ask this is because of my concern for the post-processing again - the full implications of this are not clear to me yet as I haven’t got a handle on the man hours involved.
Ultimately to my mind the most important thing we could possibly achieve in this experiment is a workflow that is efficient for all participants ‘digital~’ or ‘analog~’. From my point of view (ideally and with longer term implementation in mind) you need to be able to manipulate records quickly and on the fly in the field or very shortly into the post-ex process. The 3D digital recording process needs to be reactive and easy to implement in the field, quick to edit if things go wrong, and complementary to the existing archive and methods of recording. I remain positive about the whole thing though as I honestly believe this is what we are working towards here...
I write this not as a criticism but as a basis for discussion these are all points I have been mulling over recently and want to speak to the team about but have not been sure how to vocalise - it helps me to write them down. I appreciate that I am very much coming at it from a field archaeology perspective and do not mind being hauled up on my misunderstandings and any unfair comments :-)
That’s enough for now, next time I might try and address some of the actual archaeology… |
