The original 1975 theory was later reformulated by Abramson et al. (1978); an essentially equivalent reformulation was offered independently by Miller and Norman (1979), but, not surprisingly, credit for the reformulation has remained with the theory's originator.
The important features of the reformulation are presented in Figure 4.1. In short, Abramson et al. added Weiner's attribution theory (see Chapter 3) to the original learned helplessness approach; that is, although helplessness continued to be seen to arise from the perception of uncontrollability, the subsequent effects were now seen to depend both on the type and the importance of the event experienced, together with the explanation that the individual produced for the cause of the event. The explanatory style dimensions focused on two of Weiner's (e.g., 1972) attributional dimensions as follows: internal-external or "locus", that is, whether the cause is seen to be due to something about the individual (internal) or due to something about other people or circumstances (external); and stable-unstable (or stable-variable in Weiner's terminology), that is, whether the cause is due to something that would recur for future similar events. In addition, Abramson et al. added a further dimension, global-specific, that is, whether the cause influences only one area of the individual's life or whether it influences many areas. The net combination of these three dimensions led to the proposal that the emotional, motivational, and cognitive deficits seen in depression could be accounted for by a particular set of attributions following the occurrence of a negative event.
The crucial type of attributional style that Abramson et al. (1978) identified as a vulnerability factor for depression was if the individual made internal-stable-global
attributions (e.g., due to my personality) for the causes of negative events and external-unstable-specific attributions (e.g., due to luck) for positive events. More specifically, an internal attribution for a negative event was seen to lead to low self-esteem especially if, by social comparison, other individuals were perceived not to be helpless in such a situation (so-called "personal helplessness"). The additional stable and global attributions for negative events were seen to add to the chronicity and the generality of the deficits observed in depressed individuals.
The empirical studies that have been carried out to test the reformulated theory have been summarised, first to provide overwhelming support for the theory (Peterson & Seligman, 1984), second to fail to support the theory (Brewin, 1985; Coyne & Gotlib, 1983), and third even not to test the theory at all (Abramson, Alloy, & Metalsky, 1988; Gotlib & Abramson, 1999)! The problem centres around what is considered to be "support" for the theory; thus, questionnaire-based studies of correlations between levels of depression and type of attributional style tend to support the proposal that internal-stable-global attributions are more commonly associated with higher depression scores and, possibly, that this type of attributional style may retard recovery from depression (e.g., Hammen, Krantz, & Cochran, 1981). However, the least convincing evidence has been obtained for the proposal that individuals prone to depression have a pre-existing negative attributional style which leads to the onset of depression in the face of negative events (e.g., Coyne & Gotlib, 1983; Gotlib & Abramson, 1999). One possible problem may be the ways in which attributions are measured in many of these studies. It is increasingly clear from research in social cognition (and indeed in other domains as well) that individuals can possess "dual" attributions—explicit attributions that form the basis of responses on questionnaire measures, and implicit attributions that are evidenced more readily by suitably indirect measures (see Smith & DeCoster, 2000, for a review). It is possible therefore that depressogenic implicit attributional tendencies are not being detected by the routine questionnaire methodology used in much of this literature.
Other failures of the reformulated model include the recognition that evidence for the proposed external-specific-unstable depressive style for positive events has been extremely mixed. For example, in a meta-analysis of 104 studies Sweeney, Anderson, and Bailey (1986) reported reasonable correlations for negative events, but much weaker support for the proposed style for positive events. The theory's excessive focus on the causes of events also seems to ignore the fact that a range of other factors including the consequences of events are at least equally important in the individual's response (e.g., Hammen & Cochran, 1981). Thus, we saw in the discussion in Chapter 3 of Lazarus' (1966, 1991) notion of secondary appraisal or coping that the consequences of events can either be easily dealt with and their impact lessened or, if the individual perceives that the available coping resources are insufficient, the event may be seen to be overwhelming. Finally, although collections of hypothetical events as measured by the Attributional Style Questionnaire (Peterson et al., 1982) may demonstrate a bias towards a particular attributional pattern, real-world events, unless they are sufficiently ambiguous, may completely override an individual's characteristic attributional style. For example, the wife of the man run over by the Clapham Omnibus may well blame the bus rather than herself for her husband's death (i.e., an external attribution), she may well believe or hope that the Clapham Omnibus plays no further role in her life (i.e., an unstable attribution); nevertheless, she may still become depressed because of the loss of her husband. The links between style and emotion can therefore be weak correlational ones rather than causal ones.
Was this article helpful?