(J 962) suggested an analysis of human communication to identify and measure the relevant variables and the mIes by which they combine to achieve information transmission. The system is confined to examining encoding and decoding of messages.Coding is the process of assigning an explicit symbol or label to a referent, concept or experience. Brown (1965) and Brown & Lenneberg (1956) have shown that cultural differences exist with regard to the categorization and consequent labelling of certain perceptual experiences. Lantz (1963) experimentally demonstrated the process of encoding and decoding a set of 20 Munse11 color chips. One group of Ss chose labels to denote each chip; another group used these labels to try to identify the intended chips. A color with a high codability score was one that the first group of Ss generally transmitted accurately to the second group. In a similar experiment using geometrie shapes as stimuli Maclay & Newman (1960) examined the effects of feedback conditions (positive, negative, and zero) and the nature of the array of stimuli (homogeneous vs heterogeneous shapes) on message length. These authors indicated that differences in message length attributable to differences between ranges of stimulus arrays may be related to the codability of each stimulus in the array.The assumption on which the present study is based is that communication about impressions of people can be simiIarly studied as is communication about arrays of inanimate objects. The study examines the effects of the presence or absence of a context array of person descriptions during encoding on the efficiency of transmitting impressions of people. Efficiency is a function of both message length and decoding accuracy (Maclay, 1962). It can be inferred that presence or absence of an array of referents affects efficiency since the array is hypothesized to influence both message length and accuracy through its influence on codabiIity. The hypothesis to be tested can be stated as folIows: Impressions of people coded Psychon. Sei., 1969, Vol. 14 (2) with reference to an array of people are transmitted more efficiently than imprcssions coded independently of the array. SUBJECTS The Ss were 40 voluntecrs from introductory psychology classes and randomly assigned to one of four groups.MATERIALS AND PROCEDURES Ten hypothetical people comprising the context array wen: each represented by a person-description typed on a 5 x 8 in. card. The descriptions were drawn from earlier studies by Boyd & Jackson (\967) and Jackson (1966) and included the behavioral information characteristic of each person.Encoders wen: shown the person descrirtions and informcd that each one was an "interview summary" written by an expert who had interviewed the person scvcral times. Each encoder was then instmcted to write down a descriptive word or phrase so that a person reading their dcscription cOlild recognize which person the S intended to denote.Ten encoders received a11 10 descriptions at on ce and were encouraged to sort through the cards...