“…For example, Marlatt et al reported that alcohol consumption by men with alcohol problems was increased by instructions telling them that the beverage they drank contained alcohol but not by the actual alcohol content of the beverage they consumed, supporting an instructional set or expectancy interpretation of loss of control rather than a pharmacological interpretation. Subsequent studies have indicated, however, that actual alcohol ingestion does increase alcohol consumption independent of expectancy in severely dependent alcoholics (Ludwig, Bendfeldt, Wikler, & Cain, 1978;Stockwell, 1991;StockweU, Hodgson, Rankin, & Taylor, 1982), supporting a pharmacological basis of loss of control in those with the most serious alcohol problems. Further, Maltzman (1994) questioned the relevance of Marlatt et al's findings to Jellinek's loss of control concept because Marlatt et al's participant sample may have consisted primarily of "problem drinkers" or "alcohol abusers" rather than Jellinek's truly alcohol-dependent "gamma alcoholics"; according to Jellinek (1960), only the latter group would be expected to exhibit loss of control in response to alcohol ingestion.…”