Attribution theory holds that the affective reaction and mood that people develop in response to a situation is to a great degree dependent on what they perceive has caused the situation. Self-blame is a specific result of certain attributions and often leads to depression. The main purpose of this study is to determine if a relationship exists between a specific, character-related linguistic bias and an increased risk for, and an elevated level of, depression. This is accomplished by comparing subjects' test results in a measure of linguistic bias with the Beck Depression Inventory score and with a measurement of attributional style. Further, by drawing on the philosophical basis of cognitive therapeutic practices, it is argued that self-blame is only related to depression if it is characterological in nature and that such characterological self-blame is implicit in the linguistic style of the individual. Elevated usage of the verb "to be" in evaluating a negative life event was found to correlate with an above-average level of the somatic symptoms of depression. Subjects who preferred "to be" sentences also made more attributions of stability in regard to the hypothetical negative scenarios. Very little correlation was obtained between depression levels and depressogenic attributions. It is argued that while the usage of specific words and the application of depressogenic attributions are confounded, the use of two separate questionnaires both related to a common vignette permits some separation. While linguistic bias does not explain the development of depression, it is at least as good a correlate as attributional style. Depressogenic biases in word usage may be the conscious expression of attributional style. / Thesis / Master of Science (MS)

Language plays an indispensable role in the transmission, maintenance, and formation of culturally shared beliefs. Yet beliefs about groups, in particular, are shared despite the existence of prohibitive norms that act to inhibit their expression. This apparent incongruity suggests that cultural beliefs become shared through linguistic means other than explicit expression. In support of this, the linguistic bias paradigm proposes that linguistic bias is the implicit and unintentional expression of beliefs through the differential use of linguistic abstraction (Franco & Maass, 1996; Maass, 1999; Maass, Salvi, Arcuri, & Semin, 1989), as defined by the Linguistic Category Model (Semin & Fiedler, 1988). According to this paradigm, linguistic bias not only reveals the speakers’ privately held beliefs but also transmits these beliefs to recipients, leading to belief sharedness. The consequences of exposure to linguistic bias, however, have yet to be shown and this is the aim of the present research program. The first study focuses on belief transmission, by determining if there is a direct causal effect from linguistic abstraction to individual impression formation. Results show that biased language transmits information about individuals but the communication context, specifically whom the message is about, is also important. Given this, it is likely that the content of the message will also affect the reception of biased language. The second study thus focuses on belief maintenance, by considering the relative effects of different levels of linguistic abstraction on pre-existing beliefs. Results were inconclusive, but may have been affected by methodological limitations. The third study addresses these limitations while focusing on belief formation, by measuring the impact of biased language in the absence of pre-existing beliefs. Recipients, in general, formed beliefs that corresponded to the biased language to which they were exposed. Together, these studies suggest that linguistic bias plays a role in belief sharedness as a mechanism through which cultural beliefs are transmitted and formed. Linguistic bias, however, must be understood within the specific communication context, which also independently affects reception.