For anyone suggesting a military purpose: Look, I know that it’s tempting to always think that Romans=legions, but can’t imagine that these have anything to do with the army. The legions were stationed along the frontiers at the time these things are dated to. Here are the dodecahedra find sites: All in the west, mostly all around Gaul and Germania. So, they’re a Gallo-Roman thing, apparently. The finds that are on the frontiers can be explained by soldiers bringing them along as personal items. Also, there were usually civilian settlements around the army bases, too.
My guess is that they have something to do with the calendar or the zodiac. Twelve sides, twelve months. Maybe some kind of fortune telling device, as has been suggested. (Ancient fortune telling could get pretty weird.) A Gallo-Roman Magic 8-Ball?
The narrow range of the dating is pretty interesting. Apparently, they date from “the 2nd or 3rd centuries CE”. What does that mean, exactly? How precisely can we date these gizmos? Does it mean a spread of two centuries, or that they’re all from just around the turn of the second and third centuries? Or that we don’t know more precisely?
In any case, it seems like a pretty narrow time range by Roman standards. It almost sounds like they were some kind of fad, like an ancient Rubik’s Cube.
I think maybe you should ponder a bit more. The largest hole in the example I studied was 25mm (about an inch) in diameter. The smallest hole was just under 11mm.
That would be absurdly small for an archery target.
Most arrows would miss or at best strike a glancing blow, damaging the dodecahedron - and if you have to construct a backstop to catch the arrows that miss, you don’t need the dodecahedron as a target at all - you paint a dot on the backstop.
Thay have to be something so mundane, so utterly unremarkable in their time that they never made it into any written records, and the Romans were obsessed with record keeping. Think something ordinary like a doorstop*, not a religious or entertainment item.
*No, I don’t think they were literally doorstops, just that there were probably that common and undistinguished.
I don’t think they are candle holders. They don’t seem a very good design for that purpose. The holes aren’t going to fit every candle size and there is no reservoir to collect molten wax (although there could have been a separate dish). I would expect traces of wax to be found in the rough interiors of the dodecahedron if they were used for this purpose.
A much simpler and more efficient design for holding candles of any size would be a metal dish with a spike in it.
Wax has been found in two of about a hundred known examples and that only means that at some point in the last two thousand years someone has put a candle in them, potentially only once.
I’m not sure how it would work as a candle holder. Here’s one from Roman times - a familiar form. When the candle burns down to the bottom it will burn itself out safely in the hole. If one were to put a candle in the top hole of a Roman dodecahedron then when the candle burns down to the metal it will drop down, potentially haphazardly. If one were to fix a candle in the bottom hole then the flame will burn down inside the dodecahedron and scorch marks would be clearly visible on the interior surface.
A circular hole in a piece of metal does not seem to be a very safe way of securing a candle and ensuring it doesn’t lean to the side over time leading to a fire risk.
I was looking into the Roman dodecahedra, and I found these two videos on YouTube that suggest that they were used for knitting/crocheting. The “knobs” on the vertices apparently suggested the idea, since they resemble similar knobs on modern forms. The two videos below show people using solid-printed replicas of the dodecahedra to crochet gloves, which explains the holes (although not, perhaps, why they were of different sizes.
The hole-less icosahedron might have been used similarly, for crocheting something else.
It’s not proof, by any means, but I’m impressed that they were able to use these replicas to make fairly decent gloves.
I’m not convinced - because the diameter of a piece of french Knitting is primarily influenced by the number of pegs and the tension - the size of the hole is nearly irrelevant.
The results of the second video aren’t any different from the output of a standard french knitting process (such as mine here: https://www.youtube.com/watch?v=0LVzSuPFhNk). There is no need for a dodecahedron to make tubular glove fingers.
The first video (knitting a whole glove at once) is certainly inventive and I admire the dedication, but it’s so horribly awkward that I can’t honestly believe the Romans settled on it as a solution.
