The Roman dodecahedron is an artifact that is a copper 12-faced hollow object of Roman origin but for unknown purposes. A very mysterious thing, its purpose or even existence seems to have never been mentioned by the makers themselves. Is there a comprehensive theory with a good amount of evidence that can point us in the right direction? Or is it just something now lost to time?
These objects are indeed quite mysterious when it come to their purpose or even their overall context : made of metal (generally bronze or copper-based), they can be described as twelve pentagonal-faced, each face being opened by holes of various diameters set inside concentric circles, and whose edges are topped by small knobs. They are essentially found in the northern-western regions of the Roman empire between the IInd, IIIrd and IVth centuries CE, but are indeed not mentioned by contemporary people except in one possible monetary representation in a lugdunese as of the Ist century CE. There's little context available, because most of the artifacts were dug out without use of archeological methodology : the best we can tell is that when we can propose a context, it's might have been (but not systematically) related to votive deposits.
Several theories had been and are still being proposed so far, some more likely or unlikely than the others, but no one being decisively accepted or proven right or wrong (even if the hypothesis, once fairly popular, they might have been sort of mace heads, is essentially discarded due to the impracticability as weapons).
One of the more generally agreed on hypothesis is to see a metaphysical, philosophical or purpose : their geometric make-up is reminiscent of their role in Platonist (where the dodechedron is seen as encompassing all other perfect forms, associated with elements) and Pythagorean conceptions but possibly as well in Druidic ones. If by the time these object appear, the latter had entirely disappeared, some intellectual traditions might have survived and integrated into a religious-philosophical zeitgast the same way Gaulish deities were re-appearing in essentially Roman provincial deities; and thus both mostly centered on former Gaulish regions and expanded over militarized borders (not unlike how Epona, as a deity, was). As such, they would be associated with zodiacal and/or metaphysical cosmologic concepts.
What would be their function, however, is hard to assert : they couldn't be used as dice (as other dodecahedrons could be, for instance with this Roman dice found in Egypt) and although their similar shape (and how these dices are mentioned in late Roman sources, especially with the sorting of the number twelve) does hint at a similarity of context they did not share the same role. Alas, the lack of sources (both archeological and historical) prevents to really propose something there : it might have been an instrument of teaching astronomical or philosophical concepts but, at least in some case, possibly a decorative toll to display one's teachings, learning or philosophical/religious function especially as one found dodecahedron could have been set on top of a staff.
It had been proposed by Michael Guggenberger that their uncommonality (although not rare) and their absence of pictural or textual mentions might have been due to be objects reserved for the initiated and kept in some form of secrecy.
Other theories propose to see in them more mundane functions : multi-calibred candlesticks, precious masterworks without any other purpose than decoration, toys akin to cup-and-balls, etc. It doesn't explain why they were restricted to the regions they were found, but it might have been the case of a being part of a local cultural tradition as many existed and still exist. These wouldn't be contradictory with some metaphysical or philosophical decorum even if it wouldn't have been its main raison d'être.
It is more problematic for another recent theory proposed by Amelia Carolina Sparavigna, according which the dodecahedron was an early rangefinder, maybe for ballistics. While it can't be dismissed out of hand, there's no indications that they had been used as such : they are of importantly varied sizes and with various diameters, and have no marks of usage such as scratching or inscriptions you'd find on practical tools. That their usage was limited to northern regions is another issue there : while understandable for aforementioned possible purposes, it is less so in this context (and remains unaddressed).
At last, some theory emerged on the Internet that it would have been a knitting toll used in "cold Europe" (it's worth noting that Rhineland wasn't specifically much colder than the rest of Gaul in winter although a great deal of them were found in the region). But it have important problems of its own, namely that the manufacture of the object was complex enough and if not rare, relatively uncommon (with little more than 110 of them being found north of the Alps) and when found in a grave context at Krefeld-Gellep, was associated with wealth. You would have expected a common and practical toll to have been rather made in iron, easily available and workable by a provincial culture that mastered its use even before the Roman conquest. It doesn't make it impossible, but as being a possible rangefinder, fairly unlikely for now.
