A recent study in New Zealand validates the theory that drunken behaviour can be partially pycho-somatic. A group of students were told that they would be tested on their response to alcohol. A number of cognitive and physical tests were to be completed at various stages. The students were given drink cards to obtain drinks from the barman. Half the students were told they were being given vodka, but were in fact given tonic water, and half were given a tonic water drink with no alcohol in it.
The results were revealing - the behaviour of the students who were told they were given vodka (up to a point) were very similar. They exhibited typical drunken behaviour, but they were not drunk at all.
For those that do not understand the “Claytons” reference, it was a brand of non-alcoholic Whiskey substitute advertised in NZ with the tag “The drink you have when you aren’t having a drink” - early '80s I seem to recall
I can imagine that in a social situation where people are drinking, drunkenness can outstrip alcohol consumption by a considerable margin (feeding on the behaviour of those who are really drunk). In this situation, a rapid sobering up is indeed possible, where the drunken behaviour stops, and the clinical alcoholic effect is actually pretty minimal.
I have experienced this myself - spending a happy evening with friends and not feeling out of place because I was drinking low-alcohol beer (1-2% which was in plentiful supply in NZ - and excellent tasting as well - on thing I miss here in the UK)