The trichonopoly chain braiding thing isn’t a terrible hypothesis - the manufacture of such chain involves drawing it down through progressively smaller holes and this style of chain or braid is apparently found in a variety of decorative and utility contexts including jewellery and adornment of armour.
There are problems though - including:
No sufficiently old examples of the chain, so it’s not established that it overlaps in time with the dodecahedra.
Pulling the braid through the holes would cause very distinctive wear patterns around the edges of the holes.
The dodecahedron doesn’t even seem to be an essential part of the braiding process - it looks as though you can just knit the wire directly onto a stick.
Just to be clear, the ‘knitting chainmail’ bit in the Joe Scott video that everyone is making a massive big deal about in this thread, is a throwaway line prefixed with ‘maybe… maaaaybeee’ (and I think he probably just misspoke and conflated chainmail and metal braid).
A bigger problem - we have zero evidence that Romans used draw plates, which would be required to make the longer lengths of wire used in the technique.
I don’t think so, as he was pointing out the military context specifically.
This is an interesting theory… it fits the need for all the different sized holes, it explains the nubs, it makes sense regionally, with the Viking Weave thing, it explains why it was normally made out of bronze and copper; both harder metals than gold. It explains why they were buried with women and maybe… maaaaybeee… military guys used them to make chain mail?
Certainly that last part is incorrect, and he’s repeating existing (example) circulating misinformation and conflation of different metalworking techniques, but it’s not as though he asserts this with any kind of certainty, in fact, quite the opposite.
The earliest example of knitting is around 11th century CE.
It may be that Roman-Egyptian materials were the origin of knitting, several centuries earlier, eg 3rd to 5th century CE, but knitting as we know it would have been unknown to Roman legionaries in Northern Europe.
I was curious whether anything made using the “Viking Chain Knit” technique (aka Trichinopoly Chain) might be contemporaneous with Gallo-Roman Dodecahedrons. I started looking for examples online. In the process, I came across an article on scribd by Lora-Lynn Stevens, titled A Research Journey: Trichinopoly Chainwork Is It Viking Chain Knitting. She was trying to figure out if the technique is really old enough for Vikings to have used it. She cited books by James Graham Campbell and Leena Tomanterä that included photographs and descriptions of many pieces of Viking jewelry that appear to be the earliest items made using this technique. They date back to about as far as the 9th century CE.
It doesn’t sound like there are any examples of Gallo Roman trichinopoly chain. The technique came into existence about 500 years after the heyday of Gallo Roman Dodecahedrons. Trichinopoly chain might also require pulled wire, rather than the inconsistent wire made with earlier methods. The earliest archaeological find of a drawplate for making wire may be from Sweden during the Vendel Period, 550 to 775 CE. So, my opinion would be that dodecahedrons were not used for weaving ‘Viking Chain Knit’. I don’t think that particular technique for weaving wire had been invented yet, and the tool for making the required type of wire had not been invented yet.
Knitting in the sense of two-needle knitting, yes. But one-needle knitting (naelbinding) produces a near-identical textile, in both form and function, that only a trained expert might be able to tell wasn’t knitted, on fairly close and possibly destructive examination.
Or in other words: Romans could make warm woolen knitted socks just fine.
The knitting explanation pretty much has to be bogus - for a variety of reasons, including:
If true, this would be the only bobbin/loom knitting device of its era - that is, nobody made simpler versions that were just a thing with a hole surrounded by some pegs, or if they did, no examples of those simpler devices survived to be found. Not even one. We’re supposed to accept that, not only did loom/bobbin knitting happen 1500 years earlier than everyone thought, but they went straight to this implementation of it?
Also, every demonstration of bobbin knitting I have seen people attempting, using replica dodecahedrons, looks tremendously awkward compared to using some object like a lucet or just a piece of wood with a hole surrounded by some dowels or pegs. Even accounting for the skill and practice of the person trying to demonstrate their theory, the Roman dodecahedra would just be horrible, uncomfortable things to try to knit with.
(As mentioned upthread) the size of the hole in a knitting bobbin has almost no influence on the diameter of the knitted tube - that is determined by the count and spacing of the pegs, which is identical on all 12 faces of the objects. There’s no point in making this thing 12-sided with different holes, if it’s for French knitting.
And probably most importantly, there are examples of the object that would be impossible to use in that way - such as the smaller, more chunky versions of the object.
Yeah. Just as a comparison, this is what a Viking-era lucet (2-prong cord loom - I know you know this, @Mangetout, just for others) looks like:
compared to the sleek wooden fork used in more modern times. We should be seeing some similar progression.
It’s similar to Binford’s Curation Model, which distinguishes between curated and expedient artefacts (as a continuum, not a binary).
Under Binford’s curation model, if dodecahedrons were utilitarian:
We’d expect a spectrum from expedient to curated versions — cheap, simple ones alongside fancy ones.
We’d expect larger quantities in the archaeological record, especially from workshops, refuse pits, and everyday household/workplace contexts.
We would definitely expect local improvisations or cheaper materials (e.g. wooden or ceramic copies).
But the evidence doesn’t match:
They are always made of relatively expensive bronze.
They’re rare — only about 120-130 known across the entire Roman world (but see my point about the frequency of e.g. gladius or scutum finds).
They’re mostly found in special contexts (military sites, ritual deposits, hoards) — not everyday ones.
There’s no evidence of simpler or cheaper versions.
Since every YouTuber seems to end up making a video about these things, I suppose I will do eventually (subtitle for the video already written: ‘prepare to be disappointed’)
I was going to say, my son makes chainmail. (He is also an expert knitter of fabric.) I’m not sure if he calls making chain mail “knitting”, but i have sometimes referred to it that way, and he hasn’t corrected me. He makes chain mail as a costume component for larping, so he doesn’t use rivets, and often uses aluminum, rather than steele, rings. But either way, the rings are typically made en mass by wrapping a wire tightly around a cylinder and then cutting the coil into rings, and i bet that’s how it was done back then, too. Then you knit the rings together and individually seal each one with pliers.
Probably done that way often. Sometimes rings were stamped out of iron. The Vikings commonly attached one split ring to 4 other rings which may have been solid, either stamped out, twisted, or forge welded shut. But for any split ring using rivets turned out to be the most secure means of closing the joint. Chain mail was susceptible to breaking and even if a weapon didn’t break through the mail the broken pieces could injure the wearer. Not sure, but I think riveting rings was faster than forge welding them (heated and hammered together) because the rings could be prepared before connecting, and perhaps the tiny rivets needed no heating.
I’m sure you know that the more workable aluminum was not available to the ancients, and not very suitable against iron weapons, but aluminum would probably have been used in ancient armor to take advantage of it’s light weight if it were available. It would have been bulkier than the equivalent iron or bronze armor because it’s a softer metal, but it still would have been lighter.
ETA: I think if I was born a little later (or born younger?) I would have been into LARP like your son. I’ve met a few armorers like your son who make stuff for themselves and others. They study and apply the classic techniques in detail while adapting modern materials.
Yes, but his group (which isn’t the SCA) uses foam rubber weapons, so the aluminum chain mail works just fine, and it’s much lighter to wear, as well as easier to make. He also uses steel, and has a steel chain mail neck tie that he wears to formal occasions. I don’t actually know whether his steel or aluminum mail vests are more popular. (They back his realm’s currency, so he has to make them from time to time.)
Monty Python and the Holy Grail probably hasn’t helped - it’s somewhat widely known that the chainmail garments used as costume in the movie were actually just knitted from grey wool - it bears no relation to actual chainmail, but the concept of ‘knitting chainmail’ has entered the public consciousness in part because of this.
Does this imply that, like the joke about baby pigeons, the dodecahedrons we find likely are the simpler, cheaper versions?
That is, whatever it is these things do, they are as simple as they can be to do it and therefore every aspect of them is strictly necessary for whatever that function is?
