00:00.00 archpodnet Welcome to the show. Everyone Paul how's it going man. 00:02.83 Paul Ah, okay I think I'm ah, really really sleep deprived but but happy I guess you know came back from Saudi a week ago just over a week ago and I've got another week and a half or so before I head back? Um, yeah, so. 00:10.10 archpodnet Ah, is. 00:15.37 archpodnet Oh man. Yeah. 00:19.36 Paul We're going to have to find some more ah co-hosts for you because my schedule is awful lately? Um, how you doing Chris and where are you. 00:24.97 archpodnet Yeah, well I'm doing my yeah I'm doing all right? We're in ah north western washington if you see a temperature map of the United States we're the only blue part that is is not over ° I mean in fact, it's ° right now. The high is 72 and I am loving it I don't ever want to leave here just because of that given where everybody else is doing so yeah, ah, just so everybody knows too. We are recording video. So if you want to watch this rather than listen to it. You can go over to archeology podcast network on Youtube and check it out. So. 00:42.16 Paul Me. 00:59.46 archpodnet Let's get into the show I'm going to I'm going to introduce our guest through his bio right now and then we'll start talking to him. Okay, so Mark that for the bio. Okay, now that we know Marcus will introduce him Marcus how's it going? No problem. So. 01:12.80 Markus Perfect. Thank you so much for having me on your show. 01:14.88 Paul No thanks for coming. 01:18.31 archpodnet Yeah Marcus is fresh in from the field like literally a few hours ago as we're recording this so we pretty appreciate you coming on and and recording this interview with us. Ah, but what led you to us was an issue in advances an archeological practice that we will link to in the show notes and the title is machine learning based. Identification of Lithic Micro debotage which just sounds super cool. So I'll kick it off with a question here. Um, and we'll we'll get into a bunch of other stuff. But you mentioned right in the beginning of this article slow data and honestly I don't think that's a term I had heard before so what. 01:48.95 Markus M. 01:54.28 archpodnet What do you mean by slow data. 01:54.57 Markus You know I mean I think as archaeologists we all tend to work for a long period of period of periods of time at the same site where we really produce very deep and rich data about this specific location. That is really contextually, very important and super rich and of course where we spend a lot of time interpreting it. But from my perspective and I think many others. We also realize that it's very hard to generalize from this data set to something to patterns that are. 02:20.30 archpodnet M. 02:34.50 Markus Broader Larger where we can say oh this is something that is not only happening here locally but where we can look at patterns that are visible on a much larger scale. So and that's what and and again I mean you know I'm not critiquing slow data I mean I'm just literally coming back from the field where I'm producing. 02:44.66 archpodnet Okay. 02:53.73 Paul From. 02:53.99 Markus Ah, slow data. But what I tried to do in this paper together with my colleagues is to point out how we can complement the slow data with other approaches that include for example machine learning. 03:08.22 archpodnet Okay. 03:13.31 Paul So um, why machine learning in particular, you have an interest in it obviously but what gravit what makes you gravitate toward it. 03:23.20 Markus So as we explain in this mouthful of an article and my interest was in a lot of like understanding lithic. Ah so or that's I should say Stoneworking. So. In the area where I'm working the classic maya. They never really developed metal tools as we see it in many other complex ancient societies but they really relied on stonenapping napping off chert of Obsidian and possibly other stones. 03:37.61 Paul Um. 03:44.10 Paul Um. 03:55.20 Markus Ah, to produce all their tools and many other implements in their life and for me, it was always interesting to understand well who did this stonenapping. How did they produce their stone tools and this brought me then to the interest into Microabbotage because modern stonenappers. Tend to and get rid of their debris that they produce during stone napping simply because it's very sharp and you don't want to walk over it or you don't want your children or others are being hurt by it and so actually among the classic maya and other societies. 04:19.19 archpodnet Me. 04:21.16 Paul Um. 04:33.76 Markus We know relatively little about how the stone napping actually works where do we have workshops and 1 of the approaches that archaeologists have been toying around for decades now is the use of microdibotage meaning to look into. The microscopic debris that these stone workers are producing and the advantage of macabbotage is that in most Contexts It's very hard to get rid of them. So unlike larger flakes that you can easily pick up on surfaces microdebotage. 05:02.69 archpodnet Love. 05:10.31 Markus Gets easily lodged into floors and other areas. So this means that by analyzing Microabbotage We can get an idea about where ancient stone maps have been working and. 05:13.88 Paul Oh. 05:28.88 Markus So you get here a future archaeologist peeing in so and my my daughter and and so and so and my interest is now well to analyze Microabbotage to recognize macroabbotage in soil samples and to. 05:35.14 archpodnet No worries. 05:35.55 Paul And that's lovely. 05:48.61 Markus Find out where ancient stone Nappers have been working so in the past or should I go on a little bit about how people have done this so far or. 05:50.82 archpodnet Okay. 05:58.65 archpodnet Well, let me let me ask a ah clarifying question First I mean how do you Define Micro Debotage I read it in the article but just so we can get it out on the on the podcast here. What? how are you defining micro debotage and and how does it. How does that debotage differ than say regular Debotage. You know the tertiary primary secondary flakes. However, we identify those does it have different characteristics not just size. 06:19.00 Markus Yeah, so and this is again and actually I mean I can talk about a paper in progress that I'm just writing about exactly that. So there's a lot of discussion. You know what is microdevitage. So for this paper machine learning. We defined it as. 06:29.19 archpodnet Yeah. 06:35.41 Markus All flakes that are smaller than six point three millimeters or a quarter inch so this would be what would be seed out during regular screenings in excavations. So this is more like a horistic approach and in the paper that I'm right now writing. 06:38.59 archpodnet Okay. 06:48.24 archpodnet Huh. 06:53.58 Markus I argue that probably a tighter definition is warranted but at least for the the key ideas. Um that this this is the stuff that falls through our sifs when we are excavating and looking for artifacts now one of the other aspects that you mentioned. 07:06.47 Paul O me. 07:07.63 archpodnet Um, okay. 07:12.90 Markus This is one of these critical ideas that microabittage supposedly shows the same characteristics as recollet debitage like you know the conjoital fractures and this is again one of these issues that I can come back to where I'm where I got a little bit skeptical. 07:21.74 archpodnet Um. 07:30.29 Markus Looking at macrodeittage because for me, it's much much more variable than the regular flakes that we would pick up from a stoneworking workshop. 07:39.33 archpodnet Okay, so you yeah, go ahead. 07:40.75 Paul Um, why would that be yeah, why would that be why was there so much more variability. 07:46.33 Markus I mean 1 is simply the size. So what we did is we work with modern stonenappers. So we go to napins and for this particular paper we work with a good friend of mine Michael Mcpride he's like reproducing stone tools. 07:52.78 Paul Oh. 07:57.10 archpodnet Um, looking. 08:04.50 Markus So we asked them you know could you make a iphase or arrow hat. Whatever you want to do and can we collect your debris and most of the stone nappers that we approach. There are super happy I mean you know for them. It's just well. It's debris and and so then really care about it. But for us for my team It's really super interesting to collect all these thousands if not tens of thousands of particles and what we then do and I can come back to that we run all of these particles through a particle analyzer. So this is a machine that allows me to take photos of all of these particles and um, describe them in various ways. An interesting part to bring it back to Paul to your question when we looked at the variability. Like you know, simple dimensions. How long is each particle. How wide what is the transparency angularity. It turns out they don't fall in this clear like like a tightly defined class that we would expect but instead there is a huge variability. 09:10.76 archpodnet Um. 09:13.40 Markus And this is one of the issues that makes makes dealing with microabittage so difficult that they is that they're much more variable than people previously assumed. 09:24.37 Paul Um, and is micro debotage that you've derived from ah from human activity from Flint Napping ah is it statistically different is significantly different than other little stone bits. You might find I mean are you able to to determine that. 09:38.16 Markus So and this is exactly where it becomes really interesting because what we did for this article is we compared the experimentally produced macroabbotage to aricular soil sample that I picked up many years ago during my dissertation in Guatemala. And I should say this soil sample contained all kinds of stuff so you know like little plant twix little ceramics. Whatever was you know what you would find in a soil sample and so what we did for this article is to compare the experimentally produced macrodabotage to this. 10:16.83 Paul O. 10:17.54 Markus Other so which contains not only rocks it contains Sand Twix and whatnot and ah the interesting part is statistically I mean we applied not only the machine learning algorithms to these 2 samples but we also ran statistics and 1 of my graduate students is writing a paper about that. And we can show that these are statistically pretty different. There is some overlap. But overall if we look for specific dimensions. We can really differentiate between these 2 categories very nicely. 10:41.75 Paul Um. 10:50.48 archpodnet Okay, you mentioned when you first started getting into this that you know people have tried to study microdebotage in the past aside from, you know your approach of using you know these these tight scanning methods and machine learning to to suss this out. What have been some of the past approaches to this. 11:08.11 Markus So you know and that's what I tried about twenty years ago when I was dealing with my dissertation because I mean I was working as ah at a really interesting side in my area and my mentor encouraged me to look into soil samples and I was like well that would be really interesting because we knew. 11:12.63 archpodnet Um. 11:25.70 Markus That there was a stonenapper working at this particular location where we excavated and so I you know I used the microscope that we had in the lab and what I did was the traditional approach of analyzing soil Samples. So I pulled I think I have ah had about 50 soil samples. And I sift them into different size fractions and then I looked at each size fraction under a microscope hoping to find oh what is soil and what is microabitage and this experience I mean it was super tedious I mean literally I spent. 11:56.70 archpodnet Ah. 12:01.10 archpodnet Are. 12:01.13 Markus hours and hour hours and hour hours just looking through a microscope at soil samples and ah ultimately I mean you know I ah this experiment never made it into my dissertation because I I really questioned myself you know after looking for hours and hour hours and hour hours at the at the same soil sample I was like. 12:15.22 archpodnet Um. 12:21.40 Markus Really see that or am it just imagining that this is a tiny flake or is this really just my own imagination speaking and you know just to give you a few numbers. You know what I didn't realize at that point with the particle analyzer I mean I regularly find. 12:24.70 Paul Me. 12:25.60 archpodnet Um. 12:38.30 Markus Three hundred to five hundred Thousand particles in a soil sample. So this is literally less than a handful of soil and now you can imagine you know like the ah idea would be or the ideal for an archaeologist would be to look at each of these hundreds of thousands of particles 12:43.49 archpodnet Um, ah. 12:57.31 Markus And say oh this is soil this is microabittage and at least for me this is where I became skeptical where I said you know I mean I I can do this for perhaps one hundred two hundred particles 13:01.76 archpodnet Um. 13:11.12 Markus But to really scale it up to 500000 or even millions of particles I said you know this is I mean at least for me I couldn't say that that this was what what I could really get a quantifiable result out of that. 13:27.30 archpodnet Wow, that's really cool I'm trying to think Paul just ah, editorial note here should we go into one of those other questions or we we're getting a little close to the end of the segment. Um, do you have anything to follow up on this. 13:41.35 Paul Ah, let's ah we can go into the next the second unchecked question down after the break. 13:47.17 archpodnet Okay, okay, all right sounds good. Alright so when you're looking at this micro debotage. Um I mean the particle analyzer in it and in particular figure 1 in your paper shows I mean at this scale what it clearly looks like. Flakes when you're looking at it this scale blown up into the into the ah into this image but then also just you know soil particles that are slightly more rounded but kind of the same size. You know what? I mean um, how well is the analysis that you're doing able to differentiate between these two and. You know? Well we get into how long it took you to teach it to do that. But how how well does it actually do that. 14:27.10 Markus I I'm I'm sorry I think I just got like just half of what you said I mean on my side. The let me see what I have to. 14:34.18 archpodnet Yeah, no worries. Yeah yeah, I'll restate the question. Um, so looking at your figure 1 in the paper here you have side by side soil samples getting some audio stuff there. 14:47.33 Markus So can you still hear me I mean I can't hear you right? The the video feed is frozen. Oh. 14:49.11 Paul Now. 14:52.50 archpodnet Yeah, we can yes, okay, well why don't we go ahead and take a break and then we'll get Marcus' ' video fixed on the other side so we'll do that in the meantime check out arcpadnet.com/members. To join us for our next culture our share event and see all the other stuff that we have for you back in a minute.