As with the odor of alcohol on the breath, few police reports will fail to include an observation by the arresting officer that the arrestee exhibited "slurred speech". (See my earlier post, "Alcohol on the Breath: Evidence of DUI?"). The officer fully expects to hear slurred speech in a person he suspects is intoxicated, particularly after smelling alcohol on the breath, and we tend to "hear" what we expect to hear. And hearing it supplies the officer with corroboration of his suspicions.
Even assuming the honesty of the officer that the defendant’s speech was slurred, there is little evidence that this is symptomatic of intoxication. Impairment of speech is, for example, a common — and sober — reaction to the stress, fear and nervousness that a police investigation would be expected to engender; fatigue is another well-known cause. However, consider the following excerpt from Discover magazine:
"Bartenders, police officers and hospital workers routinely identify drunks by their slurred speech. Several investigative groups judged the captain of the Exxon Valdez oil tanker to be intoxicated based solely on the sound of his voice in his radio transmissions. But a team led by Harry Holien, a phonetician at the University of Florida, has found that even self-proclaimed experts are pretty bad at estimating people’s alcohol levels by the way they talk."
Hollien asked clinicians who treat chemical dependency, along with a group of everyday people, to listen to recordings made by volunteers when they were sober, then mildly intoxicated, legally impaired, and finally, completely smashed. Listeners consistently overestimated the drunkeness of mildly intoxicated subjects. Conversely, they underestimated the alcohol levels of those who were most inebriated. Professionals were little better at perceiving the truth than the ordinary Joes…. "
He thinks his research could encourage police to be more wary of making snap judgments: Mild drinkers might come under needless suspicion…" Saunders, "News of Science, Medicine and Technology: Straight Talk", 21(1) Discover (Oct. 2000).