Now, having said that, the majority of behaviorists and veterinarians consider a citronella collar not a punisher, nor an aversive, but rather a "disruptive stimulus." In my use of it, I have found that in about 75% to 80% of the dogs I have used a citronella collar on, it is indeed a disruptive stimulus in that they don't find it aversive, but rather they get the spray and immediately forget what they were barking about and start to investigate the intricacies of the spray. The other 20% to 25% of dogs I have used it on have indeed found it to be an aversive, as evidenced by their behavior and attitudes after the spray.
...
[FONT=Comic Sans MS, Arial, sans-serif][...]The concept of "disruptive stimulus" results from behavioral patterns. It can be defined as a jarring stimulus that interrupts the course of the sequence, which produces an expectation stage enabling enticement of the pet to another (desired) activity. The present study starts from this definition to show the usefulness of a citronella spray (device called ABOISTOP?, by DYNAVET, France) as such a stimulus in dogs showing territorial barking. 52 dogs spending at least 4 hours a day in a garden adjacent to a busy street were included. The trial compares the effects of a punitive stimulus consisting of a garden-hose spraying water, to the ABOISTOP? collar. Treatment was allocated at random. Once a week, each owner noted the frequency of barking towards 10 pedestrians. A first control was done on day 0 (which provides the reference figure), then every seventh day till day 35 when therapy was stopped. Relapses were to be assessed on day 90. It should be noted that every bark interruption by the device was immediately followed by a play session initiated by the owner (redirection of behavior). [emphasis inserted by Doug] In the "punishment group" we could note a sudden cessation of barking, as early as on day 7, which was subsequently steady until day 28. In the "disruptive group" the decrease of barking was more gradual (48.6% barking on day 7 - 16.9% on day 14) and a total disappearance could be heard by day 21. In addition, the relapse rate on day 90 was 86% in the "punishment group" versus 3.8% in the "disruptive group".