Semantic Satiation of Words and Numbers*

Abstract

One purpose of the study reported is to compare the effect of continuous verbal repetition on the semantic ratings of words and numbers. A second is to examine the effectiveness of the semantic differential to index distinctive features of numbers as well as words. The procedure used is parallel to an earlier study on semantic satiation with words with the notable difference that, in the present case, a completely mechanized method is developed to present stimuli and record rating responses and their latencies. The following main findings are reported: (1) Regular semantic differential wales reflect significant satiation effects with words, but not with numbers, while a ‘Meaningful‐Meaningless’ scale indexes significant shifts toward the ‘Meaningless’ pole with numbers, but not with words. (2) Certain differences are observed between words and numbers in the manner in which these are rated on standard semantic differential scales. (3) The M‐shaped relationship between response latency and scale position reported earlier by Osgood is confirmed in the present paper. The implications of these findings in relation to the earlier study and with respect to the applicability of the semantic differential to numbers are discussed.

Initial polarity, semantic differential scale, meaningfulness, and subjects' associative fluency in semantic satiation and generation.

“…Repetition of the words failed to produce satiation or generation, either overall or at any level or combination of levels of initial polarity, scales, m, and 5V associative fluency. These re-sults are contrary to those reported by Lambert, Jakobovits, and their collaborators (e.g., Lambert & Jakobovits, 1960;Messer et al, 1964) and others (Amster, 1964). But, as noted before, others have also failed to obtain semantic satiation or semantic generation (e.g., Floyd, 1962;Reynierse & Barch, 1963;Schulz et al, 1965;Yelen & Schulz, 1963).…”

“…Lambert, Jakobovits, and their collaborators (e.g., Lambert & Jakobovits, 1960;Messer, Jakobovits, Kanungo, & Lambert, 1964) and others (Amster, 1964) report the occurrence of semantic satiation or generation of the meaning of words. These phenomena are defined, respectively, as a decrease or an increase from their initial semantic differential (SD) ratings which is attributed to interpolated repetition of the words.…”

Review of the measurement of semantic satiation.

“…Though more formal experimentation is necessary on this point, informal observation by one of the present writers does find a tendency of subjects to add all three numbers. Messer et al (1964) found no differences between the initial and the final latencies of semantic ratings for numbers, indicating that their satiation group did not differ from their control groups with regard to response latencies. Neither did they find evidence for number satiation from the semantic ratings.…”

“…Reynierse and Barch (1963) found no satiation effects, and Schwartz and Novick (1964) concluded from their study that the satiation treatment may have nonspecific, nonsemantic effects. Using both numbers and words for stimuli, Messer et al (1964) found no satiation effects for numbers on the standard semantic differential scales, but did find an effect on a special meaningful-meaningless scale. For words, the satiation group showed a satiation effect on the standard semantic differential scales, but, unexpectedly, so did one of their control groups.…”

Some methodological issues in semantic differential research.

“…While lack of a refined metric may not interfere with most uses of the SD, this inadequacy could create problems in certain kinds of work. For example, studies of semantic satiation (e.g., Messer, Jakobovits, Kanungo, & Lambert, 1964) postulate that when words are repeated continuously, they become meaningless, and therefore, ratings on SD scales should drift toward center. Tests of this idea depend directly on the assumption that the true zero point of each scale is located at the middle check position.…”