The long ramp down of my summer research leave is starting to give way to the gradual ramp up to the fall semester, but before I let the entire summer slip away, I’m making sure I have the time to enjoy as much music as possible.

After a winter and spring filled with Miles Davis (and looking forward to a summer and fall filled with Coltrane), I decided to listen to a couple of my other favorite trumpeters as a palate cleanser. A couple of weeks ago, I enjoyed some Blue Mitchell (with Stanley Turrentine). This week, I listened to two live shows in Baltimore.

The first one, Live in Baltimore 1968, features Lee Morgan and Clifford Jordan along with John Hicks on piano, Reggie Workman on bass, and Ed Blackwell on drums. That’s a formidable group and this is a formidable album.

https://youtu.be/ZCTr-4U7_LA?si=uewve8eWB4A4r6RD   

Staying in Baltimore and with lead trumpet, I put on Understanding, Roy Brooks 1970 date at the Left Bank Jazz Society. This raucous live album featuring Woody Shaw on trumpet, Carlos Garnett on sax, Harold Mabern on piano, and Cecil McBee on bass.

Understanding was the second track on the A side of another live album, The Free Slave, recorded that the same year. It features George Coleman on sax and Hugh Lawson on piano. I don’t know Lawson at all, but he’s more restrained than Mabern giving both Shaw and the underrated Coleman more space to work. It’s remarkable to think that Shaw is only 26 on this album. McBee, on both albums, does McBee things which are magical. 

If you’ve not listened to this album, oh man, you should. It’s tighter than Understanding, but lacks none of the propulsive energy.

All three of these albums constitute what my friend Sharon Carson would call: “get out of the house music” which is appropriate for a Monday morning!

In some ways there seems to be natural line from slow archaeology to film photography. Maybe the time that it takes to get film developed (and the expense!) nudges us to slow down and be more deliberate with how we make photographs. On the other hand, the instant results from digital photography do encourage a kind of iterative practice that maybe the time and cost of film photography doesn’t quite support (or protracts in certain ways).

All this is to say that I took some film photos for the first time in 20+ years. I shot Kodak Tri-X for the black and white and Kodak Gold for the color. I used an almost 20 year old Nikon N90S with a Takumar 55 f1.8 lens (that didn’t allow for infinity focus) and an Olympus Pen EE-2 half-frame film camera. And, I do look at the back almost very time that I release the shutter. 

Right now, both cameras are loaded with some expired Kodak Max 400. I have a hunch with what I’ll get, but we’ll see in a week or so.

For now, fresh film. First from the Olympus:

I’m still pretty intimidated by film, so I’ve been reluctant to load it into my more serious Nikon. In the little Olympus, shooting half-frame feels pretty close to digital: low stakes, lots of shots, and all I have to do is wave the camera in the general direction of something interesting and hope for the best. In the Nikon, there’s control and precision and any mistakes are impossible to disown.

But here are some of my favorite from the roll including one of the last photos that I took of the late Mighty Milo. 

I spent the last few days alternating between reading, thinking, and muddling around with NDQ using Claude Code. About 50% of my time involves cleaning up the data so that I can do more sophisticated analysis with this text. To get a sense for what I’ve been working on, check out the post on Tuesday. 

One of the oddities that I encountered with the first analysis was The Bellman clustered by itself with little overlap with any of the other contemporary magazines in the corpus (NDQ, the two NDSHS journals, DeLestry’s Western Magazine, and The Midland). Interrogating this further, I discovered that it represented a larger problem. Apparently the model I was running was only reading part of the text (the first 2000 or so words) to create the model. This was not acceptable and a good reminder to verify then trust.

To address this, we devised a plan to break the entire 20 million word corpus into around 100k, 200 word segments and to perform a BERT analysis on all of these segments. As part of these we also continued to remove stray text — such as headings— that exerted a surprising influence over the analysis. It looks like this is going to be an ongoing problem moving forward because it crops up from time to time, but continued efforts will mitigate it over time. In any event, once we implemented this new method, The Bellman was brought back into the fold and our BERT topics became a bit more clear. The UMAP scatter looks a good bit different from what the analyses that I produced on Tuesday appeared.

The Bellman is reintegrated into the model and in some ways parallels the diversity of North Dakota Quarterly across the entire scatter sharing space with both The Midland (and my impulsive inclusion of some of Hamlin Garland’s work in dark and light orange) and the NDSHS journals (in pink).

The next step was to model similarities across the magazines and here I faced a new challenge. The Bellman and NDQ had a whopping 55% overlap and The Bellman and The Midland has a 66% overlap! This seemed exceptionally large and defied what I know of the two magazines. 

What we discovered is that the Bellman represents nearly half of the total corpus of 200 word segments. As a result, it is statistically more likely to have similar topics with all magazines owing largely to its numerical preponderance.To compensate for this we developed a “Chance-adjusted lift” that compensated for numerical prevalence of The Bellman (and the small number of 200-word chunks from, say, The Midland). This would temper the prevalence of The Bellman and boost similarities between Garland and The Midland which show clear topical affinities.

This is all well and good, but it assumes a kind of regional affinity as opposed to constructs it. (Ideally, I would include a non-regional magazine here to demonstrate that the connections between these publications are not simply mapping across the topics so broad that any publication might fit; this is something that a subsequent analysis would resolve). Below are the top 8 topics (including the almost random “mrs, miss, mr mrs, mr” category which speaks to The Bellman’s interest in Minneapolis society more than anything else). 

It is notable that The Bellman and NDSHS publications are the most distinctive voices in terms of topics and The Midland is the least distinct. In other words, a chunk is most likely be assigned to a particular publication on the basis of topic and style. Garland is funny for his <50% level of confusion meaning his texts are less likely to be assigned to Garland than elsewhere!

Again, this still doesn’t really speak to any particular regional character (although some of the terms in the topic list do speak to this). A more focused regional analysis had to relay on certain key terms associated with these publications. These terms ranged from “frontier,” “prairie,” “pioneer” and “our region,” “our section,” to “Dakota,” “the Northwest,” and “Minnesota,” and places such as “Grand Forks” and “Minot.” This allowed us to identify explicitly regional discussions in the chunks and to characterize these discussions. One thing that I had not necessarily named was the “boosterism” that characterized DeLestry’s Western Magazine (and to a certain extent The Bellman, NDQ, and The Midland). 

Most interesting is that NDQ has elements of each of these types of articles and this seems more or less stable over time and irrespective of editor and by extension genre. Later editors of NDQ included more poetry, for example, but this didn’t necessarily shift the percentages at the volume level. The parallels between the character of publications in NDQ and their topics (in the chart above) shows just how eclectic NDQ was and how its various elements had parallels with other contemporary regional magazines while mimicking none of them.

The final analysis that I worked on this week was understanding how certain kind of sentimental language present in elegiac and personal memoirs (which so frequently characterize regional narratives [i.e. the experience of the region]) suffuses other discourses from boosterism to antiquarian and even scientific and literary (which here is literary criticism) works. I developed this by looking at how the more elegiac passages talked about the region. Then I looked across the more ostensibly prosaic expressions of regional identity to determine whether this vocabulary appeared when discussing pioneers, the frontier, our section or individual places. Unremarkably, some of this language did appear in more prosaic passages: especially in the boosterish and promotional texts and more obviously (and self-referentially) in the elegiac-personal memoir texts. On the other hand, it is unsurprising that scientific and technical texts would not indulge in sentimental language other than, perhaps, in a critical and demythologizing vein. It is perhaps a bit more interesting that articles interested in imagining the nation through abstract concepts (race, “the country,” or census data) sough to demythologize the some of the conventional sentiments around the frontier, the pioneer, or the prairie. Pieces with literary and aesthetic concerns (e.g. namely reviews, but not exclusively), recognize the picturesque and romantic as characteristic of much of the sentimental literature that it reviews.

This analysis was based on 1200 pieces (there is a challenge with the text of The Bellman and DeLestry’s Western Magazine that they can’t alway be aggregated into articles). The next step, of course, is for me to go through these passages and assess and even refine these categories. 

The goal of this analysis wasn’t to replace a careful reading of these magazines, but to help navigate the over 20 million words of text captured in these issues. In particular, I was interested in discerning how the region not only manifests in these issues, but also how these periodicals constructed the region. The next steps in this process is to clean up the archive of python scripts, texts, and analyses. Then, subject it all to a critical eye. And finally, return to my history of NDQ and determine how and whether this supports my analysis.  

I am trying to work my way thought my summer reading list starting with the top of the pile. This means reading some articles from the latest issue of Hesperia.

I was particularly intrigued by Nikos Magouliotis article “The Printed Page and the Painted Column: An Architectural Microhistory of a Church in Ottoman Thessaly.” The article considers a 19th-century crossed-domed basilica in the village of Palamas on the Thessalian plain. The church itself, while attractive, is hardly remarkable except perhaps for its dome and rather impressive size.

Magouliotis article doesn’t so much focus on the architecture, although he notes that the construction of a church of this size in the Late Ottoman period demonstrates the tolerance of Christianity under Ali Pasha’s reign in Central Greece. Magouliotis, however, studies 19th century decoration of the church with particular attention to the painting on the south wall of the nave and the decoration of the columns. 

The south wall depicts a series of heart shape figures which are unusual in Byzantine iconography. Evidently they depict the stages of the heart of man. This wall painting dates to 1892 and the iconography derives from an early 19th century book by the German Catholic priest (and later convert) Johannes Gossner. Published in Berlin in 1812 and then in subsequent editions and English translation as late as 1851. Somehow this book made its way in the hands of an artist in a village in Thessaly. 

This book does not appear in the small library of largely 18th century liturgical books found in the church’s library. These books, mostly published in Greek in Venice, largely lacked illustration and were not books that prescribed church architecture or decoration. Instead, these books featured Baroque ornament that complemented the sincere and pious texts. Magouliotis revealed, however, that these textual ornaments attracted the curiosity of 19th century readers who sketched and doodled copies of the small scrolls and flourishes in the margins and on blank pages. These doodles may have been the basis for the Rococo decoration on the church’s columns.

For Magouliotis, this connection offers a small window in the history of this otherwise typical church’s decoration. By doing this, Magouliotis complicates the notion post-Byzantine wall painting by demonstrating that the circulation of books did not produce slavish dedication to either “Western” or “Byzantine” influences, but rather creative appropriation of elements of emerging book culture in the Southern Balkans. The adaptation of decorative elements from books to the architecture of the church is distinctive and offers a window into the creative life of this particular community. What is significant and telling is that this doesn’t map clearly onto a trajectory that assumes provincial communities were seeking to actively align themselves with Western Europeans styles and culture. 

This kind of microhistorical analysis — à la the late Carlo Ginsburg — demonstrates how individual decisions made on the very local level constitute agency in the transformation of culture at the more collective (and abstract) scale of the community or “society.” This modest effort to return agency to the residents of a small village in Thessaly is lovely tribute to Ginzburg who passed away this year. I like to think that by publishing this article, the Hesperia editors (and the author) offer a small (micro?) tribute to the passing of one of our generation’s greatest historians.

N.B. Some key parts of this post have been superseded by my post on Thursday. Check it out.

Over the last couple of weeks, I’ve been working on figuring out how to do some remote reading of NDQ. I’ve done this largely through using Claude Code to deploy its robots to read both the first 23 volumes of NDQ but also the output of other contemporary magazines from the ND State Historical Society, The Midland (a literary magazine), and The Bellman and De Lestry’s Western Magazine (a weekly magazine published in Minneapolis) and the work of Hamlin Garland (for various reasons associated with the emergence of late-19th and early-20th century regionalism).  

First, I’m not very familiar with remote reading (and this branch of digital humanities more broadly). What I’m doing here is mostly flailing around. To be clear, the paper that I’m writing based on this analysis already done. In fact, it’s already too long, so most of this analysis is just to see where it leads. 

Second, it’s proven a bit tricky to get the text extracted from the various journals in a consistent way. De Lestry’s and The Bellman, for example, are in columns and more like newspapers meaning that an article can sometimes appear across a number of non-sequential pages or columns (depending on the placement of advertising). This means that the sentence level integrity of what’s on the page is preserved but it is detached from the article level integrity.

The main kind of analysis that I’ve conducted is BERT (Bidirectional Encoder Representations from Transformers). This allowed me to come up with a list of 65 topics developed from the BERT analysis. On review of this, there are some obvious problems. For example, “The Midland” appears as a running head on pages of the journal and this has “corrupted” (in a sense) the BERT analysis. This same problem may appear in The Bellman. Even as I was writing this, I was re-running the analysis to try to fix this kind of thing.

Even allowing that this has influenced the BERT analysis, certain patterns have emerged that are consistent with my less systematic reading of the content of these magazines. Below is a UMAP (Uniform Manifold Approximation and Projection) graph of the five journals and Hamlin Garland’s work. The lightest blue is NDQ. 

(This has been heavily revised and my new analysis appeared on Thursday).

First, it appears that NDQ is an expansively hybrid journal that touches on topics of that appear in most of the other publications. Some coincidences are not surprising: NDQ has clearly shared interest with the two journals from the State Historical Society of North Dakota. This is owing to their shared interest in North Dakota as much as shared authors (especially Orin G. Libby and his students). The overlap between The Midland and NDQ is most significant in areas of literary criticism. 

Second, The Bellman, a prominent Minneapolis magazine, is the most different from the other magazines. Curiously, The Bellman has at its nearest neighbor in terms of content with De Lestry’s Western Magazine with over 90% overlap and nearly no nearest neighbors with any other magazine. NDQ, in contrast, has content that has neighbors with almost every other magazine (other than The Bellman) with its closest neighbor being The Midland. It’s important to note that this represents a quantitative study of the topics and context based on their numeric fingerprints. It’s a heat map of content, but an unsubtle one.

One thing that became clear here is there ND Historical Quarterly (NDHQ) and ND Historical Society Collections (NDHSC) are functionally so similar that they are the same thing. They do not overlap chronologically. The resulting content overlap with NDQ remains strong and slightly more clear than the interlocking content of The Midland and Hamlin Garland’s work and Western Magazine and NDHSC/NDHQ. 

The goal of this analysis — such as one existed for as wide open a project as this — is to trace the construction of a regional discourse across a range of periodicals. It is interesting to note that The Bellman, with its focus on Minneapolis society, contributes least to the broader regional conversation. In contrast, Western Magazine, despite its close connection to The Bellman and its Minneapolis roots, shares topics and content across NDQ and the ND state historical society journals. The other axis of regional content follows Garland’s 19th and early 20th century writing, The Midland, and NDQ with greater connections to the historical magazines of the ND state historical society than the Minneapolis magazines. 

Finally, my hope is to make all this data available with my article. Even if I mainly rely on human reading of the texts, this kind of remote, digitally mediated reading, has the potential to generate patterns that can inform new hypotheses that we can explore either computationally or through human reading of content.

More on this as I continue analysis…

For some reason, when I put on music in the summer, I don’t want something that is too demanding, but I do want something that is good. This often draws me to soul jazz playlist, particularly the mid- to late-1960s offerings of Blue Note (and when I feel a bit more polished, CTI). 

Yesterday, for example, I enjoyed Dr. Lonnie Smith’s 1969 album, which has a fantastic cover and features Smith (on organ, of course), Lee Morgan, Julian Priester, guitarist Melvin Sparks (who I think brings a grittier, more late-1960s guitar to Grant Green lines), Bennie Maupin (!!), and Leo Morris on drums. With Smith on the organ, no bass is required. There’s a good bit of relaxed improvising on this album over deep grooves.

I’ve started to get into Stanley Turrentine (if you’re going to listen to summertime soul jazz, you’re going to listen to Turrentine). I hadn’t listened to his 1970 date with CTI, Sugar. Turrentine is joined by George Benson, Freddie Hubbard, Lonnie Liston Smith (not the Dr.) on electric piano, Ron Carter, and Billy Kaye on drums (as well as various other performers including Hubert Law and Airto). It’s a very chill album that never quite drifts into the background. To me this is the essence of soul jazz. It’s relaxed, but it’s never shallow or background music (and when it does flirt with smooth jazz, it’s time to find something better to listen to). Its the kind of music that you put on while you drink a Negroni or a gin and tonic and enjoy.

For some reason I hesitate to call Ike Quebec’s Heavy Soul (1962) soul jazz. It has an organ (Freddie Roach), which for some makes it soul jazz by definition, and Quebec has plenty of big tenor sax soul in his sound and maybe that’s enough. But its early 1960s date sets apart from the sound of mid- to late-60s commercial soul jazz recordings (and, yeah, I know cats like Lou Donaldson were recording soul jazz albums in the early 1960s). Maybe it’s Quebec’s breathy phrasing that makes me feel a kind of 1950s spirit or his interest in ballads framed by Roach’s unctuous organ that makes me think of supper clubs and long summer evenings.

I’ve been taking more film photos which I’ve just sent off to get developed and that’s cut into my photo Friday habit. So, today’s will be a bit more of a grab bag.

A few Sony photos from our last days at Polis on Cyprus:

Some with the Sigma DP2 Quattro:

Some with the Ricoh GR3:

Some with the Nikon D600:

And, finally, some of my attempts to capture the smoke from the Canadian wildfires:

Over the last few weeks, I’ve been playing with Claude Code and building little applications that solve very particular problems. A few people have asked me what I’ve been using Claude Code for and how it’s gone. 

Last week, I built a little semantic search application for the first 23 volumes of North Dakota Quarterly. This was in the service of an article for the Edinburgh Companion to the Regional Magazine. I talked about what I did in last week’s Three Things Thursday. What is remarkable to me is that it has worked consistently enough to be a real help in my research (which is largely done, but I’ve benefited from the ease with which I’ve been able to cross check my conclusions against articles from the Quarterly). 

The goal is eventually process all the volumes from NDQ so that I have a searchable database of all articles from 90+ volumes. This will take some time (and some tokens), but now that the prototype is built, all I need to do is feed it the scans. In a perfect world, I can make this available publicly (somehow).

This make me think it was a good time for a Three Things Thursday focused on some of tinkering.

Thing the First

Yesterday, I was faced with the unpleasant task of removing the background from 300+ photographs of painted plaster and molded gypsum from PKAP and adding publishable photographic scales. As readers of this blog know, I’m all for slow archaeology and finding something positive and useful in the tedious work associated with archaeology, publishing, and academic life. Removing the background of photographs, however, during the summertime, when I have many pressing responsibilities (including, say, walking the dog) feels like an unnecessary burden. 

This past summer, Richard Rothaus showed me how he got Claude Code to remove the background from artifact photos and told it to not touch any pixels associated with the artifact itself. He also gave it instructions to modify the scale meter into something publishable. I applied most of what he showed me to the plaster and gypsum photographs and added that I wanted to have a viewer that would allow me easily compare the original to the modified photo.

Thing the Second

Over the last ten years or so, I’ve been puttering around on a project that tries to map the 1910-1950 Grand Forks Census data. This project is more complicated than I anticipated owing to changes in street names, lost neighborhoods, and challenges cleaning up the data. That said, I’ve made some slow progress and added some of the lost neighborhoods to a GIS map and tried to manually clean up the data. It was pretty tedious work, but it taught me something about how the census worked.

At this point, I’m mostly hunting for outliers and this is no longer edifying. More than that, there is the complicated task of placing addresses on the map that no longer exist in the contemporary grid. These are things that our robot friends are particularly good at doing. They can follow patterns and normalize data, identify outliers in large datasets.

Since we’re giving a paper on Cold War Grand Forks, having a better sense for the organization of the city in the 1950s will be a useful thing!

Thing the Third 

One of the looming challenges at Polis is area EG0. We’v not started any serious study of the buildings there, but we know that the first step is to read and process the excavation notebooks. At Isthmia we’ve experimented with using robots to transcribe the notebooks and creating facing page transcriptions. This seems like a good model for Polis notebooks as well.

Polis notebooks are also complicated because they do not follow a last-in-first out order nor did excavators limit themselves to one context open at a time. As a result, excavated contexts are scattered across notebook pages and usually I manually had to arrange these contexts so I could read the excavation of a context continuously while still being able to understand its place within the greater narrative of work in the trench.

As we turn to studying the excavations of E.G0, we need to transcribe and rearrange these notebooks and this seems like a good use case for the robots. I’m keen to get working on this in the spring.

In May, I was asked to write a book review of Anna Magdalena Blomley’s 2022 book, A Landscape of Conflict? Rural Fortification in the Argolid (400-146 BC). I hadn’t written a book review for a couple of years and has assumed that I was out of the rotation (there seems in my life to be an inverse relationship between the number of book reviews I am asked to write and the number of manuscript reviews I’m asked to perform. It was pleasant enough to review Blomley’s book which I think I read in 2023 or 2024 and was able to re-read in the service of this review. The book is good and most times a solid book is harder to review than a remarkable or deeply flawed one. 

Here’s my first effort at a review. It still needs a few sentences at the end and a careful proofread, but otherwise it is more or less ready to go.

Anna Magdalena Blomley’s A Landscape of Conflict? Rural Fortification in the Argolid (400-146 BC) (Archaeopress 2022) demonstrates the potential and limitations of applying a landscape approach to the fortifications of the Argolid. Blomley admits that this effort was only partly successful owing as much to the difficulties dating fortification as the incomplete character of our understanding the Argive landscape. That said, she does succeed in putting the an expansive assemblage of rural fortification “on the map” through her extensive discussion of the Argive landscape and catalogue of 146 fortified sites. As importantly, Blomley’s work offers a thoughtful critique of arguments for the function and organization of rural fortification by locating the sites in a regional perspective.

The book is divided into eight chapters with the first being an introduction and the last a conclusion. The four chapters after the introduction provide a basic survey of the history and topography of the Argolid (Chapter 2) which will offer nothing of particular note to anyone familiar with the region. Chapter 3 defines a typology of fortifications in the region and distinguishes between towers, forts, fortified settlements, and drystone circuits. The next chapter confronts the “persistent problem” of dating these fortifications. Literary sources offer dates for only two fortifications (a 5th century fort near Palaia Epidavros and 2nd century diateichisma at Troizina) another hand-full of fortifications, such as the Larissa at Argos, the Archaic circuit at Asini, and fortifications associated with excavated sites with Porto Cheli in the Argive Akte, being datable through excavation. To supplement the tiny number of archaeologically datable fortifications, Blomley offers “local masonry chronology” that provides provisional dates for 94 of the 146 sites in her catalogue, although these dates are quite broad (e.g. Late Classical to Hellenistic) compared even to archaeological chronologies.

Chapters 5, 6, and 7 situate the fortifications in a regional context and this is most significant contribution of this volume. Chapter 5 summarizes the current state of knowledge on settlement and territorial borders in the Argolid. The former drew upon the limited evidence from surveys in the Southern Argolid, the Berbati valley, and the Nemea Valley. Blomley also summarizes what is known about the borders of Argos proper and the dependent urban centers such as Kleonai, Orneai, Nauplio, Mycene, Tiryns, and other settlements brought under Argive control during the Archaic and Classical period. This provides a foundation for Chapter 6 which considers the role of rural fortifications in territorial defense. Blomley used GIS models of least-cost-path and viewshed analysis to challenge ideas that rural fortifications functioned as part of defensive systems connected through lines of sight or designed to monitor or defend important routes between or through territories. She demonstrates that there is no evidence for a territorial defense network that encompassed the entirety of the Argolid. Notwithstanding the challenges of dating the fortifications, she argued that the limit evidence that did exist involved only a small percentage of the rural fortifications, were likely regional in character, and represented responses to local topography and geopolitics. Chapter 7 argues that some rural fortifications would have functioned to serve the needs of local communities. They may have offered places of refuge and protected agricultural processing equipment such as olive presses. None of the features associated with local use during times of duress, however, precluded other functions — from housing garrisons to monitoring traffic — and many likely served multiple functions over time. The indeterminant function of rural fortifications suggests that many represented decentralized strategies determined by local conditions and situations rather than centralized strategies instigated by Argos. Imposing viewsheds and lines of sight and the proximity to known or possible routes and borders may have been epiphenomenal to the construction of forts and towers in locations advantageous to local needs.

The strength of this work rests in the combination of over 140 known fortifications of the Classical and Hellenistic periods with a bewildering range of evidence for settlement, rural activity, and topography. The critical use of agricultural catchments, least-cost-path analysis, walking distances, and viewsheds and intervisibility took advantage of the proliferation of high-quality digital elevation models designed to facilitate regional level study. The legible (and colorful) maps included in the volume complement and support Blomley’s arguments. The number of sites and the geographic scope of here analysis limit how much even the best drawn maps can help the reader keep track of the bewildering array of site names, regions, and landscapes. In fact, to replicate (and to critique fully) Blomley’s analysis, one would have to painstakingly extract the maps and the catalogue from book. This would be a time consuming and imprecise process, but perhaps worth the time and energy in light of the persistent scholarly interest both in the Argolid and in regional fortifications.

Last week, I was invited to a conference in September at the Center for Hellenic Studies in Washington, DC. They need a title by the end of this week and I think that I have a title and some kind of basic idea for a paper: Slow Archaeology and Punk Practice Revisited: In Praise of Fragments.

On my walk the other day, I start to spitball some ideas and here is my very preliminary outline:

1. Slow Archaeology and Punk Archaeology Revisited (~6-7 min)

The organizers told me that they were interested in my ideas of slow and punk archaeology. I have to admit that my original formulation of these ideas (as well as the author himself) is getting a bit long in the tooth these days. Fortunately, some thoughtful people have picked up these ideas and refined them. There’s also been a good bit of parallel work done across the humanities, social sciences, and even academia in general. While this might be too self-indulgent (even for me), I wouldn’t mind presenting both a basic framework of slow archaeology and a review of how some of the ideas have  developed over the past decade. In particular, I want to inject a discussion of how generative AI might serve as a kind of provocation that can challenge and complicate notions of slow in archaeological practice.

2. Punk Practice, Conviviality, and Publishing (~6-7 min)

The second part of the paper will pick up with David Pettegrew’s Corinthian Countrysides: Linked Open Data and Analysis from the Eastern Korinthia Archaeological Survey (2025) and our recent article in the Journal of Field Archaeology. My interest here is to continue to develop Ivan Illich’s concept of conviviality as a bridge between slow archaeology and punk archaeology. In particular here, I want to focus on how conviviality plays out in the work of publishing itself and demonstrate how digital tools when combined with new forms of community (both scholarly and otherwise) can take precedence over the pull toward ever increasing levels and forms of efficiency.

3. In Praise of Fragments (~6-7 min)

This is where I want to go off the rails a bit. I want to suggest that one of the ways that slow practice can subvert the drive for efficiency is by celebrating the fragment. In particular, I want to bring in Byung-Chul Han (and to a lesser extent Walter Benjamin) to argue that fragments are not subordinate to narrative which is common sense in archaeology. After all, archaeology begins with fragments: they pre-exist narrative and as a result can never be fully subordinated to it. The fragment resists synthesis rather than waiting to be completed by it. Han, in particular, makes clear that fragments resist incorporation into narratives and resist efficiency. For Benjamin, this was a good thing because narratives were so rife for abuse (especially for the Fascists); for Han fragments subverted efficiency by becoming objects of contemplation and reflection. Here, if I’m feeling it, I might return to the role of AI in archaeology. Generative AI seeks “narrative consistency” as a tool to unify fragments. This isn’t always benign. In conversation-based generation, a fragment that doesn’t fit the narrative context isn’t reconciled; it’s discarded. This makes AI structurally counter-fragmentary, not misused technology but a clear expression of digital technology’s long-running drive to smooth over anomaly (GIS, digital photography, the compact disc, generative AI).

I’m not sure where this all goes in terms of real world applications. I like to imagine that by spending more time with unresolved fragments we create a way to find Han’s contemplative space and cultivate a refusal of urgency, both academic and broader-cultural. Whether this approach can apply to the digital analogues of archaeological fragments is hard to know.

Needless to say, this will require more thought before it’s ready. I have a couple of months to work on it, though.