Joe Roe

Assistant professor of the archaeology of the Middle East at the University of Copenhagen. I use data science and ecological modelling to investigate prehistoric agriculture in West Asia.

Also https://archaeo.social admin, #Rstats developer. I usually post in English, nogle gange også på dansk.

Website: https://joeroe.io
ORCID: https://orcid.org/0000-0002-1011-1244
GitHub: https://github.com/joeroe

Public Key

npub1yyhpl57vehe0e0cygfe94wgvt52mcpa6cdc3nulr9vxhva77kz9qc7fk7p

NIP-05 Address

[email protected]

Profile Code

nprofile1qqszztsl60xvmuhuhuzyyuj6hyx969duq7avxuge703jkrtkwl0tpzsprpmhxue69uhhyetvv9ujumt0d4hhxarj9ecxjmnt0g3xj3

Publishing to

relay.momostr.pink

Author Public Key

npub1yyhpl57vehe0e0cygfe94wgvt52mcpa6cdc3nulr9vxhva77kz9qc7fk7p

Show more details

Last Notes

2026-02-09 08:43:16 GMT

‘AltMetric’-style services, which track citations to academic papers outside of academic literature—so in the news, social media, etc.—have been quick to incorporate #BlueSky posts, but I’ve yet to see one that includes the #fediverse
I can see that (real) decentralisation makes it harder to find citations to papers in the Fediverse, but it ought to be possible.
Is there anybody out there working on some sort of FediMetric service?
#AcademicPublishing #Science #Publishing

2026-02-04 16:30:33 GMT

- reply

I remember talking to my (Danish) wife about this. When we first met she thought I used singular 'they' because I was very progressive. Actually it just comes naturally to me as a native English speaker for people whose gender I don't know, and from there it's not a big step to extend it to nonbinary people.
Now I have to remind myself that you can't do the same in Danish. Constructions like "Kunden har altid ret, har han ikke?" sound as weird to me as singular they does to you.

2026-02-02 08:19:56 GMT

- reply

The obvious solution is to go out with chisels and disrobe the fellas.

2026-01-29 08:31:55 GMT

Bitter vetch might have been food after all, @npub17x4…u7yh
https://www.tandfonline.com/doi/abs/10.1080/14614103.2025.2610045
#Archaeology #Food

2026-01-29 08:09:53 GMT

- reply

The end of the month deadlines are upon me so what I *should* be doing today is finishing marking last semester's papers, finishing preparing next semester's teaching, finishing the edits to a manuscript my coauthor asked for, asking for permission to apply for a grant to do fieldwork next winter...
What I *have* been doing is wrestling the toddler into his winter clothes and then getting distracted by the idea of using geospatial AI to find and classify desert archaeological sites.

2026-01-23 15:35:43 GMT

- reply

Sounds like the machine whipped up something that uses #Rstats under the hood, since that's a common R error message!

2026-01-22 09:52:45 GMT

- reply

Amazing what massive government subsidies and systematic sabotage of the competition can do!

2026-01-07 04:34:32 GMT

- reply

If you take the Vancouver rules, for example, I can't see why a student or hired excavator isn't considered to have "contributed substantially to [...] the acquisition [...] of data for the work". Honestly I don't include them as coauthors either—I've never heard of a project that does—but I couldn't give a reason other than "tradition in the field".
This only applies to the first publication of excavation data, of course – you don't get put on every paper that uses it.

2026-01-06 19:41:56 GMT

- reply

Is archaeology still fitting the humanities trend in that regard? The mean number of authors on my papers is 6, for example.
I've always thought that we're quite stingy with authorship. Routinely excluding the people who physically excavated a site from the primary publication of it, for example... it's not like the research could have been done without them!

2026-01-01 10:13:14 GMT

- reply

And makes them no less effective bartenders after graduation! Just more indebted and dissatisfied ones.

2026-01-01 10:11:49 GMT

- reply

My university recently announced that external funding will be de-emphasised in hiring and promotion from now on (part of a long-running controversy in which the university claims grants end up costing them money), but let's see if that's what actually happens in practice.

2026-01-01 09:50:24 GMT

- reply

Plus the vast majority of humanities PhD positions today have nothing to do with reproducing the discipline, they're a byproduct of getting Big Grants on their supervisors' CVs. Because the Big Grant system was designed for natural science and the only way we can spend remotely comparable sums of money is with junior researcher salaries. But that's a whole other discussion.

2026-01-01 09:43:23 GMT

- reply

It's a good point that would go down like a lead balloon with most of my colleagues. To have a realistic discussion about the humanities job market within universities, we have to remove the existential dread. We fight tooth and nail for education in Niche Studies because if there's no students, there's no positions. If society acknowledged that it's good to have expertise in a niche even if there's no jobs in it, most humanities profs would happily go back to one student a decade.

2026-01-01 08:59:05 GMT

- reply

Entrenched snobbiness distorts the graduate job market in the UK. British Widgets Co. would prefer to hire someone who studies Latvian poetry from Oxford than one in Applied Widgetometry from Oxford Brookes. 'Transferable skills' might be the fig leaf invoked but really it's about maintaining elite access to good jobs, exacerbated by the overexpansion of universities.
Scandinavia has fewer universities with less pronounced differences in prestige, so this doesn't happen.

2025-11-07 22:46:56 GMT

- reply

I won't post the actual list because it includes people with follower-only accounts, but—measured by the number of posts in the last six months—you are indeed the talkiest user here, posting about six times more frequently than the next runner up.
So thank you for keeping things lively! :) The stats show that we have a lot of active users who (like me) read often but rarely post.

2025-10-17 14:58:46 GMT

- reply

Honestly not a bad description of early colonial archaeology though…

2025-10-09 08:25:10 GMT

- reply

Ooh that's fantastic. I've also heard that a lot of that data for Denmark is available, it's just not been put together and/or made public. Hopefully the new version of F&F will take some lessons from across the Öresund!

2025-10-09 07:25:27 GMT

- reply

I’ve always been impressed by Fund og Fortidsminder. Haven’t looked at the Swedish equivalent, though — what extra info does it give?

2025-10-02 05:58:54 GMT

Can anyone point me to literature on static websites in open science/open data? I feel like the advantages are obvious and I'm sure I've read bits here and there, but I'm struggling to find a good reference for a grant application...
#staticweb #smallweb #indieweb #OpenData #OpenScience

2025-09-30 11:51:14 GMT

- reply

It's a funny one. But I suppose we have to bear in mind that at that time, the other options available (cereals, etc.) weren't nearly as palatable as they are today. People were used to having to heavily process plant foods to get a decent meal.
It also wasn't actually one of the first to be cultivated or domesticated (https://doi.org/10.1007/s00334-023-00917-1). It comes in relatively late; late enough that there probably were domestic cattle around to feed.

2025-09-26 15:01:11 GMT

- reply

One at least should be seen as a positive mental health development!

2025-09-05 08:27:11 GMT

- reply

It's very sad. When I visit contract excavations (in Europe) I'm always blown away by the amount of manpower, resources, and tech they can bring to them. Meanwhile in research excavations (in the Middle East) we're squeezing what we can out of shoestring budgets or else writing yet another paper on a site excavated 50+ years ago. It's quite a mismatch.

2025-08-31 07:21:48 GMT

- reply

Many of the smaller Danish islands have been depopulated in the last fifty years. They tend to be bought up by the state or charitable funds for nature reserves or by rich people for luxury resorts. Withdrawal of public services speeds things along...

2025-08-21 08:54:28 GMT

- reply

There are also great FOSS options there: Krita instead of Photoshop for digital painting; QGIS instead of ArcGIS for maps; Inkscape instead of Illustrator for vector graphics.

2025-08-17 11:45:14 GMT

- reply

I wonder if this is why self-conscious 'paradigm shifts' in archaeology dried up after the 90s. Young scholars from Binford's and Hodder's day were in a position to stick around and follow through on their manifestos, later generations weren't?

2025-08-04 14:01:06 GMT

- reply

Especially when that 'pattern' is only seen at <insert famously weird site here>!

2025-07-17 06:56:26 GMT

- reply

Err, more like 15 mins :)

2025-04-15 10:27:18 GMT

- reply

Yes exactly! This would let you estimate the 'population' of finds in the ploughzone as a whole and therefore what fraction of it you recover in any given survey. It's mark and recapture, essentially (https://en.wikipedia.org/wiki/Mark_and_recapture). We rarely get to use it in archaeology because we rarely put things back.

2025-04-15 08:16:42 GMT

- reply

Do you mark them in some way? Would be cool to do that then try surveying the same spot again in a few years, see how many marked objects you get and so estimate your detection rate.

2025-04-03 06:52:09 GMT

- reply

Well, not me... I can barely tell the difference between a flake and a blade. But yes in general; they are Natufian and Pre-Pottery Neolithic. We have one site (Shubayqa 1) that has been fully excavated and mostly published, then several more that are in progress.
Some summaries:
https://www.degruyter.com/document/doi/10.1515/9781789201574-029/pdf
https://www.journals.uchicago.edu/doi/abs/10.5615/neareastarch.80.2.0094?casa_token=boVag_LCEmkAAAAA:A79g04CNMi6NzIkkes3MYFBEm5Fof3htQzL9qbRhWn2FimGL0FD80SOl81Y4F7kfWgWwtYLSusSQ
https://researchprofiles.ku.dk/da/publications/first-inhabitants-the-early-prehistory-of-north-east-jordan

2025-04-03 06:43:34 GMT

- reply

I think so too! It's a good point of comparison to Europe, where you rarely get such clear signals of climate impact – shows that when a society is *really* 'climate-sensitive', it's not hard to spot.
Shubayqa in eastern Jordan. Hopefully the paper with this analysis will be out before the end of the year.

2025-04-03 06:40:02 GMT

- reply

Well it's Palaeolithic sites around a transitory desert wetland, so more like dry climate = no people. And it can be hot dry or cold dry, depending on what the precipitation is doing.

2025-04-03 06:31:55 GMT

- reply

I finally got my last paper off my desk earlier this week, so now I can focus on the data analysis for the next one – which is always my favourite part. It's still a work in progress but it looks like the palaeoclimate data lines up pretty much perfectly with the presence/absence of archaeology, which is nice, for once!

2025-03-19 10:24:42 GMT

https://archaeo.social has defederated https://social.freysa.ai, which is flooding hashtags such as #Archaeology and #History with repetitive LLM-generated posts.
#MetaArchaeoSocial #MastoAdmin

2025-03-13 07:41:25 GMT

- reply

He didn't preserve any of the surface of the headstones, as far as I know, so they were effectively destroyed. I think it is that or gravel, like Sweden, unfortunately.

2025-03-13 07:31:27 GMT

- reply

In Switzerland I had a neighbour who was a sculptor. He'd buy old headstones in bulk – they were a cheap source of good stone for smaller pieces.
In eastern Jordan we tend to see settlements and monuments from successive places being built on top of each other, right back to the Palaeolithic, but I suspect that's more to do with each layer creating a more and more prominent topographic feature in an otherwise flat landscape than an understanding of them as 'heritage'.

2025-03-07 10:01:59 GMT

- reply

Sad. But they've been trailing this for a while by making letters absurdly expensive. It cost us thousands of krone to send our wedding invitations out last year!
I suppose there's nothing stopping you putting a letter in a box and sending it as a parcel...

2025-02-26 15:50:30 GMT

- reply

I think there's an interview somewhere or other where Sid Meier said he based the first Civ on his high school world history textbook, which was clearly a product of 40s/50s American social (neo)evolutionism.
I love the games but in terms of their theory of history they're an absolute horror show: evolutionist, essentialist, environmentally determinist, techno-utopian... etc. etc.

2025-02-26 14:39:11 GMT

- reply

Oh no, we understood it all too well – Raichu is clearly not a tale of progress.

2025-02-18 18:48:50 GMT

- reply

The main use seems to be writing things that people don't want to write, which is funny because the output is something that people don't want to read.

2025-02-04 12:52:40 GMT

- reply

The single biggest expense we have is storage of media files attached to Mastodon posts. To reduce these costs, I'm considering the following options. It would be very helpful to know which (if any) you would be okay with.
NB. This is just to gauge what people think – there's no concrete plans to do any right now. Assuming income doesn't improve (see above), there are other technical measures we could try to reduce storage costs.
#MastoAdmin

2025-02-04 12:47:44 GMT

- reply

We're not in any immediate danger but clearly we need to adjust the model. I have a few ideas and would love to hear more.
There are three ways you could help to raise our income right now:
1. Contribute financially at https://opencollective.com/archaeosocial
2. Consider whether your lab/society/etc. could use our #Matrix server (https://about.archaeo.social/services/) for its internal communications and build a regular contribution into its budget
3. If you know of any grants that we might be eligible for, let me know.

2025-02-04 12:47:08 GMT

So there's a bit of a difficult conversation to be had about this instance's finances. #MetaArchaeoSocial
From the start we've run entirely on a collective, pay-what-you-can model through https://opencollective.com/archaeosocial.
Since our operating costs are modest (c. €1 per active user per month), my hope was we could be sustainable this way. Two years in, it is clear this is not happening. The gap between costs and contributions is widening and we are now running a €190 per month deficit.

2025-02-04 07:36:10 GMT

- reply

There are some (e.g. https://fedidb.org/network/instance/archaeo.social) - but they're at a more aggregate level. You could get a lot more detail from the API (https://docs.joinmastodon.org/api/guidelines/), but I don't know of a tool that compiles user-level stats automatically.
We don't collect info on where people are accessing the site from (for good reason, I think) but the most active languages last month were English (67% of posts), French (11%), German (11%), Spanish (6%), Italian (5%), Dutch (2%), and Polish (1%).

2025-01-31 09:06:27 GMT

- reply

Much appreciated :)

2024-12-31 21:43:14 GMT

- reply

Did you see https://employment.ku.dk/faculty/?show=163014 ?

2024-12-13 14:30:18 GMT

- reply

That is... hard to believe!

2024-11-20 08:18:45 GMT

- reply

Interesting, I've never though about assessing the probability that way around before. It would be quite straightforward to do it with a date calibrated in R or Python (like the one in the image). For example, in R:
Ua75650 <- c14::c14_calibrate(1116, 30)
Ua75650[[1]] |>
dplyr::filter(age <= 1060, age > 970) |> # 890–980 AD
dplyr::pull(pdens) |>
sum()
#> [1] 0.8014018
So 80.1%
OxCal is great software, but people do tend to treat it as a black box...

2024-11-19 07:11:36 GMT

- reply

I don't know if it's so much a difference of era but a difference between HE in Scandinavia vs. HE in the Anglosphere. "Graduate jobs" that don't depend on a specific degree are just not as big a thing here (in Denmark), which combined with more specialised and less flexible degree programmes make it much harder to advise students they should approach education for its own sake, even if I agree with it in theory...

2024-09-19 07:30:55 GMT

Can't satisfy #GitHub's increasingly invasive requirements for verifying academic affiliation for free access to features like private repositories ("Take a picture of (not upload) a dated ID" / "Why aren't you on campus?" / "Upload an image explaining why you're not on campus") so I guess I'm paying $4 a month for GitHub Pro now. Is this a win for Microsoft?
I really need to get around to moving to #forgejo...

2024-09-04 09:56:58 GMT

- reply

I think it's when the action involves multiple actions and elements. You "prepare" a meal when you don't just cook it, but also cut up the ingredients and put it on a plate. You "prepare" a figure when you don't just draw it, but also make measurements and format it for printing.
"Made" doesn't quite work as a substitute, but I can't put my finger on why – maybe because it implies something more physical?

2024-08-03 09:19:32 GMT

- reply

Best I've managed is a couple of sneaky Star Trek (https://joeroe.io/posters/2019/caa2019_qcontinuum.html) and Firefly (https://joeroe.io/talks/2019/caa2019_fieldwalkr.html) references :)

2024-08-03 08:21:33 GMT

- reply

Have you tried the reverse?

2023-04-12 15:10:17 GMT

What did the first farmers actually #farm?
Eight 'founder crops'—emmer wheat, einkorn wheat, barley, lentil, pea, chickpea, bitter vetch, and flax—have long been thought to have been the bedrock of #Neolithic economies.
But early #prehistoric sites in Southwest Asia keep turning up more 'exotic' (to us) species, like the sedge tubers mixed into the earliest known breads (https://doi.org/10.1073/pnas.1801071115) or the 'new glume wheat' (https://doi.org/10.1016/j.jas.2020.105258), an important ancient crop that's now extinct.
https://mstdn.archaeo.social/media_attachments/files/110/186/473/580/639/963/original/4c0281b448a219f6.png