A tag is the name a plant gives to a thing it measures. On an estate of any size the question is not what any one tag is called but whether every system agrees — because supervisory control, the historian and the reporting layer each hold their own view of the same plant, and they are only describing one estate if they are pointing at one hierarchy.

Where they are not, a point added at one plant is intelligible only there. Every other system needs a private map to reach it, the maps are maintained by different people, and they drift. Nothing announces this. The estate keeps working and quietly stops being answerable: every system reports correctly and no question that spans two of them can be asked.

Naming is the cheapest governance available and the hardest to retrofit. Agreeing a convention before the second system exists costs a document. Reconciling two conventions after the fact costs reading every tag by hand, on a live plant, with no way to prove the mapping is complete.

The vocabulary arrived about two decades after the work did. What is now called ISA-95 asset modeling, and latterly the unified namespace, describes the same idea: one addressable hierarchy of the physical estate that every system refers back to rather than each keeping its own. The reasoning did not need the name — the same way the control and corporate separation on that estate was a Purdue-model IT/OT boundary before anyone called it one.