In all cases though, about the only way to "just stop that" is to kill them, and the messages they send each other are radically varied and diverse.
And there's a perfectly viable set of stories to explain how this all happened...
1) A bacterium that reacts differently according to the density of population of its sister bacterium is an advantage. (Work as an individual when relatively alone, work as a film when dominating the environment, etc.)
2) The easiest chemical signal to give some kind of density are chemicals that the her sisters are already expelling (doesn't need to spontaneously create a new signal-molecule creation mechanism).
3) The subtleties of "are those
my sisters or some of my cousins' sisters" is useful to know, once local competition becomes such that consobricide may be necessary, but not at the expense of sororicide, so ever more subtle hints of compounds unique to direct relations become key triggers in different branches of the cellular tree.
And so, for millions of years, these mechanisms have refined themselves (including refining themselves to make change and adaptation of the first-order event ever easier), through a drunkard's-walk of all possible biochemical processes, so that working solutions thrive and beget more working solutions, and the best we can do is to go in heavy-handed and identify the broad brush-strokes of such mechanisms and occasionally drop a whole truckload of anvils on the vague area of the small nut we're actually trying to crack. It's almost a miracle we've got
any handhold on the situation. Certainly there's no panacea, nor a carefully-fashioned nutcracker for each and every
type of nut we're encountering.
One day, however, just maybe there'll be more advanced methods of fingerprinting a bacterium's signals and synthesising something (either as a blocker or a 'fake safe' that equally messes up the process)