The Nature of Negative Thoughts in Depression



Negative perceptions and feelings about the self dominate the thoughts of the depressed. Recent theories (Abramson, Seligman, & Teasdale, 1978; Beck, 1967; Seligman, 1975) have focused on the relationship of these negative cognitions to depression and, more generally, on the interaction between mood and cognition (Bower, 1981; Isen, Shalker, Clark, & Karp, 1978). In these theories, cognitions assume a major role in maintaining a depressed mood. The depressed, in contrast to the nondepressed, perceive themselves negatively (e.g., Beck, 1967; Lewinsohn, Mischel, Chaplin, & Barton, 1980); accept more personal responsibility for failure (e.g., Kuiper, 1978; Seligman, Abramson, Semmel, & von Baeyer, 1979); and remember less positive and more negative information about themselves (e.g., Derry & Kuiper, 1981; Nelson & Craighead, 1977; Teasdale, Taylor, & Fogarty, 1980). Beck’s (1967) cognitive theory of depression provides a framework for these findings.

Beck has proposed that maladaptive cognitive schemas are responsible for the self-derogating thought patterns of the depressed. Cognitive schemas guide the selection and interpretation of information from the environment. The cognitive schemas of the depressed are be lieved to consist of negative thoughts about the self, the world, and the future. Kuiper and his colleagues (Kuiper, Derry, & MacDonald, 1982; Kuiper, Olinger, & MacDonald, in press) have suggested that the negative schemas of the depressed are primarily negative self-schemas that consist of generalizations about one’s own lack of worth, ability, or competence and the sad or unhappy feelings that accompany these thoughts.

The views of both Beck and Kuiper suggest that the depressives’ negative cognitive set determines which information is perceived, how it is interpreted, and how easily it will be recalled. There are, however, a wide variety of questions still to be answered about the specific nature and content of depressed thought. The purpose of the present study is to inquire into the precise nature of the cognitive schemas of the depressed by analyzing the content and organization of the negative thoughts accompanying depression. The literature on depression suggests that the social experiences of the depressed are often quite negative in nature and hence accompanied by feelings of anxiety and withdrawal (cf. Coyne, 1976). We assume here that the negative schemas of the depressed are generated in large part as a consequence of observing and evaluating their behavior in these social events and situations.

Moreover, we assume that these negative schemas have a pervasive impact on all aspects of thinking. To evaluate these assumptions, several straightforward empirical questions can be asked about the nature and content of depressed thought. First, is the negativity in depressives’ thoughts confined to the self, or does it generalize to others as well? Second, are depressives’ thoughts about social events and situations more negative, and perhaps less differentiated, than those characterizing the nondepressed? And, finally, what types of cognitive tasks or activities are most likely to reveal negativity in the thoughts of the depressed? The three questions posed here are central to the relationship between thinking and depression and arise from several different sources. The first question concerning the relationship between thoughts about the self and others comes from Beck’s (1967) suggestion that the depressed contrast themselves unfavorably with others.

One implication of this view is that depressives’ thoughts and inferences about themselves should be quite different from their thoughts and inferences about others. Some empirical work suggests, in fact, that although the depressed do not differ from the nondepressed in their thoughts about other people, the two groups do differ in their thoughts about themselves (Garber & Hollon, 1980; Kuiper et al., 1982; Lobitz & Post, 1979; Sweeney, Shaeffer, & Golin, 1982). This work, however, has focused primarily on depressives’ expectations and attributions (with the exception of the Kuiper et al. study, which examined recall). The present study explores depressives’ thoughts about themselves and others in a broad range of tasks to determine the generality of this finding.

The second question posed here arises from the finding that interpersonal or social skills are impaired during depression. The depressed are more likely to experience rejection in social situations (Hammen & Peters, 1978; Howes, & Hokanson, 1979), respond inappropriately to others (Hokanson, Sacco, Blumberg, & Landrum, 1980), and receive less pleasure and more discomfort from social interactions (Youngren & Lewinsohn, 1980). These findings suggest that the depressed may have somewhat fewer social interactions and that those in which they do engage will be relatively more negative. The nature of their social experiences is likely to be reflected in their thoughts about social situations in general.

Finally, the third question asked in this study attempts to determine the extent to which depression influences a range of cognitive activity. That is, will a bias toward negativity be evident throughout the cognitive continuum, or is it confined to certain types of cognitive activity? We thus examined depressed thought across a range of cognitive tasks that involved imaging, prediction, inference, and recall. To investigate these three questions about the nature of the negative thoughts accompanying depression, mildly depressed and nondepressed students were presented with a variety of social situations and events and asked to form images, to recall, and to make inferences about the events. These events were sad or happy and either social or nonsocial in nature.

While performing these cognitive tasks, some subjects were asked to focus on themselves, and other subjects were asked to focus on another person. Subjects were instructed to imagine either themselves or another person in each event and to rate how vividly they were able to imagine the event and the likelihood that the event would happen. To explore the impact of negativity during retrieval, subjects were tested for their recall of the events. The influence of negative thoughts on inferences was examined by having subjects estimate the likelihood of several possible interpretations of the events, ranging from positive to negative, either for themselves or for their friend.


The stimuli consisted of 60 slides of sentences that described different events. The sentences included a wide range of events to ensure that all subjects would categorize some of the events as sad and some as happy. The subject’s perception of the event as happy or sad was critical for our purposes. Therefore, the subject’s rating of the event as happy or sad was used to determine the affective tone of the event. Some events that were likely to be perceived as neutral were included so that subjects were not thinking only about highly emotional events. The events also varied in whether they were social or nonsocial in nature. The event was classified as social if the target person interacted with or was directly affected by another person and nonsocial if another person was not directly implicated.

After completing the BDI, subjects viewed 60 sentences that described different events. Before viewing the slides, subjects were given either the self-referent or otherreferent instructions. In the self-referent condition, subjects were asked to focus on themselves and imagine and make predictions about themselves in a variety of events. Later subjects recalled these events and made inferences about themselves from the events. In the other-referent condition, subjects were asked to describe a relatively new, samesex friend and imagine and make predictions about their friend in a variety of events. Later subjects recalled these events and made inferences about their friend from the events. Subjects were instructed to select a friend whom they had known for less than 4 months and who was not a best friend.

Previous research suggests that the familiarity of a target influences the processing of information about oneself and others (Keenan & Bailett, 1980; Kuiper & Rogers. 1979). As the target becomes more familiar, the differences in processing information about oneself and others decrease. Therefore, we asked each subject to select an acquaintance, rather than a best friend, to ensure that the subject clearly distinguished between herself and her friend. After writing a brief description of their friend, subjects participated in the same tasks as in the self-referent condition, using their friend as the referent.

A sentence was scored as correctly recalled if the meaning of the recalled sentence or sentence fragment captured the meaning of the original sentence. In a first analysis, recall was computed as the percentage of sad events recalled of the total number of events each subject rated as sad and the percentage of happy events recalled of the total number of events each subject rated as happy. In this analysis, the depression variable did not significantly influence recall by itself or in interaction with any of the other variables. Analyzing the events with this type of idiographic approach, however, severely limits the number of events that can be included in the recall analysis. That is, the initial set of events is reduced to those events rated as sad or happy and then further reduced to the sad or happy events that are recalled.

Thus, a second analysis also was performed that included all of the events that were recalled. In this analysis, it was necessary to use our classifications of the event’s affective tone. In general, subjects’ classifications of the events agreed with our classifications. In the second analysis, recall was computed as the percentage recalled of the total number of events for each type of content. An analysis of variance including depression; self-other condition; the event’s affective tone (positive, negative, or neutral); and social content indicated a reliable interaction between depression and social content, F(l,41) = 5.41,p< .03. No other effects including the depression variable were significant.

Author: R. Pietromonaco  , Hazel Markus