Children who build acquire a specific vocabulary, and adults can accelerate that considerably by using precise terms rather than simplified ones. It costs nothing and it is one of the better-evidenced contributions an adult can make.
Adults tend to simplify when talking to children, using general words where specific ones exist. That is well intentioned and unhelpful, because children are entirely capable of learning precise vocabulary and generally enjoy it.
Specific words also enable more precise thought. A child who knows the word diagonal can describe and therefore consider something they could otherwise only gesture at.
Base, frame, brace, span, joint, support, load, balance, stable, collapse.
These arise naturally in building conversation and each names something a child has already encountered physically. Naming what they have experienced is considerably more effective than introducing a term first.
Systems where the structure is visible rather than hidden make this easier, since the frame and its bracing are exposed. Rod-and-spool builds such as tinkertoy models show their own geometry, so words like span and brace have something to point at.
Axle, pivot, lever, fulcrum, rotate, friction, clearance, gear.
These come up as soon as a child builds anything that moves, and each explains something frustrating: a wheel that binds is friction, a wobbly arm needs a pivot rather than a fixed joint.
In use, during the activity, without ceremony. Handing over a piece and calling it the long rod. Noticing that the diagonal is what stopped it collapsing. Wondering whether there is enough clearance for the wheel.
What does not work is introducing a word and then testing it. Children detect that immediately and it converts play into assessment.
Children absorb vocabulary through repeated encounter rather than through instruction. Using a word consistently over weeks does more than explaining it once.
They will also use it wrongly for a period, which is normal. Recasting, using the word correctly in your reply without correcting them, is more effective than pointing out the error.
Spatial vocabulary appears repeatedly in research as a factor in spatial development, which is itself associated with later performance in mathematics and science.
That is a reasonable argument for taking it seriously, though the more immediate benefit is simply that a child who can describe what they are building can think about it more clearly.