Hmmm. As I thought I illustrated, I'm not sure that's true in relation to hypothetical scenarios such as eric presented (much akin to the metal bath one). IF, as eric was doing, a designer wishes to contemplate (and attempt to minimise the danger of) the remote possibility of a piece of floating (i.e. not earthed) metal coming in contact with the live filament of a broken lamp/bulb in a toppled over standard lamp, creating a high pd between that floating metalwork and any (earthed) exposed-c-ps, then the only way in which (s)he can take steps to reduce the danger of that (very improbable) scenario is to connect the (floating, not earthed) metalwork in question through a bit of G/Y cable to some part of the installation's earthing system.
Suffice to say that I think Eric's hypothetical scenario is fatuous.
Should his lamp fall regularly then he may consider
earthing the pipe desirable to disconnect the supply. He may then consider
bonding other pipes to the first to reduce pd between them.
However, securing or disposal of the lamp would be a much more satisfactory procedure.
The designer may think of that bit of G/Y as minimising the pd which can exist between the (not earthed) pipes/radiators and (earthed) exposed-c-ps - in which case I would call that 'bonding', wouldn't you?
There would be no pd therefore the first connection would be earthing.
If the pipes are not earthed they cannot be bonded to a exposed-c-p.
(if not, what is your definition of 'bonding'?).
A protective conductor providing equipotential bonding - where have I seen that before?
On the other hand, the designer may think of that bit of G/Y as creating a situation in which eric's theoretical event would result in a protective device (OPD or RCD) operating, thereby removing the potential danger - and I think we would all agree that is 'earthing'.
Yes, the first connection to an otherwise isolated part.
However, there is no way to 'tell' the G/Y to achieve one of those things but not the other - which is why I think that what one calls the function of the conductor is pretty moot, and that, in that situation, the 'two functions' of that the G/Y are pretty 'inseparable'.
That, obviously, is the result but it does not detract form the original reason for its implementation.
If, as in my description above, the first
earthing conductor were removed the remaining
bonding conductors would not provide an
earthing function.
Take an example of a worrier deciding to isolate all the copper pipe in his home by inserting plastic sections at all the relevant positions (entry, boiler, immersion etc.).
Eric buys the property.
Would it be wise for him to earth his isolated visible central heating pipes because of his wobbly lamp?
I don't expect you to agree