I . . . am not convinced that this is accurate. At least, I'm not convinced that it explains why "pants" is still plural. The original construction does explain why they would initially have been referred to in the plural, but it doesn't explain why the plural usage would persist to our day. E.g., in French, there is an archaic form, pantalons, but the modern usage is the singular pantalon. So why didn't we switch over the more "logical" singular form in English, as well?According to several costume historians who have helped me with this reply, the answer to all this conventional plurality is very simple. Before the days of modern tailoring, such garments, whether underwear or outerwear, were indeed made in two parts, one for each leg. The pieces were put on each leg separately and then wrapped and tied or belted at the waist (just like cowboys’ chaps). The plural usage persisted out of habit even after the garments had become physically one piece. However, a shirt was a single piece of cloth, so it was always singular.
I'd argue that it's because a pair of pants is still characterized by the two pant legs which are largely separate, even if fastened together at one end. Along those same lines, we say a pair of tweezers, pliers, tongs, scissors, handcuffs, (sun)glasses, binoculars, and goggles, and you can't argue that each of those once existed in a form where the two parts were actually physically separate from each other.
I don't fault the costumers or anyone along the line for passing along this information, because it's a reasonable-sounding explanation, but it does strike me as more of a folk etymology.
) and ask a pointed question like "A constructionist approach would suggest that this is "pair" because of other similar constructions." And then Tally M (or some other linguist? Not sure if there are any others right now) could research construction grammar for a month and tear her hair out.