Our results suggest that motivational deficits in depression may stem partly from a reduction in the specificity of personal goal representations and related cognitions that support goal-directed behaviour. Importantly, the findings have the potential to inform the ongoing development of psychotherapeutic approaches in the treatment of depression.

Overgeneralization has been investigated across many domains of cognitive functioning in major depression, including the imagination of future events. However, it is unknown whether this phenomenon extends to representations of personal goals, which are important in structuring long-term behaviour and providing meaning in life. Furthermore, it is not clear whether depressed individuals provide less specific explanations for and against goal attainment.

Introduction

Depression has long been known to be associated with negatively valenced thought content [1]–[2], but investigators have also noted that depression is associated with a tendency towards abstraction and overgeneralization, particularly for negative thought content relating to the self [1], [3]–[5]. Aaron Beck first noted the tendency of depressed patients to overgeneralize single negative events by thinking about their meaning in global terms, often with pessimistic implications for the person [1]. For example, after being involved in a minor traffic accident, a depressed person may conclude that they are completely irresponsible, reckless and blameworthy. Here, we investigate for the first time whether depressed patients' tendency to overgeneralise extends to idiographic motivational constructs in the form of personal goals and the explanations provided for, and against, goal attainment. Because reduced motivation is a principal symptom of depression [6], understanding the specificity of the cognitions relating to personal goals may prove helpful in determining the nature of such deficits.

Early cognitive theorists commented on depressed patients' tendency to think in monolithically negative terms about themselves and to provide overgeneralised, global attributions for negative life events [2]. The tendency to make general (and personal and stable) attributions for uncontrollable events (as in the previous example) is considered to be a vulnerability factor for depression according to the reformulated learned helplessness model and hopelessness theory [7]–[8]. Empirical research indicates that depressed patients do indeed have overgeneral self-attitudes and make global, self-deprecating attributions for negative events [3]–[4]. Depressed persons are also more likely than non-depressed controls to overgeneralize the implications of false negative feedback in laboratory studies [9].

The tendency of depressed persons to overgeneralise extends to the representation of their past selves. Thus, in response to retrieval cues, depressed persons have difficulty recalling memories of specific personal events that took place on a particular day (e.g., ‘the first time I visited the Eiffel Tower’), and instead tend to report broad categories of events that abstract across several episodic memories (e.g., ‘holidays’), relative to controls [5], [10]. This tendency to retrieve ‘overgeneral’ autobiographical memories occurs for both positive and negative retrieval cues, predicts maintenance of depression prospectively, and persists after remission from depressive episode [11]. Furthermore, overgeneral memory recall interacts with stressful life events to predict depressive symptoms in students [12]. A developmental strategy of affect regulation, deficits of executive control, and a tendency to ruminate on abstract themes have all been proposed as possible explanations for the phenomenon, but consensus is lacking on which of these accounts best fits the data [13].

Increasing research suggests that the cognitive processes involved in reconstructing detailed episodic memories are also implicated in the generation of future representations of events involving the self [14]. Consistent with this, there is evidence that the tendency of depressed people to report overgeneral self-representations extends to the imagination of future events that may occur to them. Thus, while depressed individuals generate fewer examples of positive future events (e.g., ‘my best friend's wedding’) than do non-depressed controls [15]–[16], depressed and suicidal people also tend to report future events that are lacking in specific detail, compared to controls [17]. This overgenerality effect for past and future events has also been shown in dysphoric adults [18].

One particularly important representation of the future self is the personal goal construct. Personal goals (e.g., ‘to pass my driving test’) have been defined as internal representations of desired states [19], which are important in organising long-term behaviour and providing meaning in life [20]. Given that episodic details of imagined future selves are more readily generated in the context of goal-related knowledge [21]–[22], the reduced specificity of future events in depression may also be manifested in more abstract representations of personal goals (e.g., ‘gain skills’ rather than ‘obtain an advanced diploma in metalwork’). Although dysregulated goal pursuit has been implicated in the aetiology and maintenance of depression [23], a recent review [24] explicitly suggests that reduced specificity of goal representations is a key marker in clinical depression. However, to date, no research has tested this latter assumption.

Although goal specificity has not previously been investigated by comparing clinically depressed and control groups, undergraduates who described their goal strivings in more abstract (or overgeneral) terms reported more depressive symptoms than those who described their goals in more concrete (or specific) terms [25]. Another study in a non-clinical population found that the personal goals of depressed and mixed anxious-depressed adolescents were less specific than those of non-depressed individuals [26]. Interestingly, this specificity deficit emerged on approach goals representing desirable outcomes (e.g., ‘always be popular’) and on avoidance goals representing undesirable outcomes (e.g., ‘avoid becoming unfit’ [27]). These results, which suggest that reduced specificity is not limited to goal content of a particular valence, mirror findings that overgeneral memories are equally prevalent across positive and negative events [11].

A few studies have examined the specificity of goals in clinical populations. Recent suicide attempters generated less specific goals, but not fewer goals, than did hospital controls, although the depression status of these individuals was not known [28]. Recently, Crane and colleagues [29] found that suicidally depressed persons showed increases in goal specificity after mindfulness-based cognitive therapy, compared to persons receiving treatment as usual, and that reduced goal specificity was significantly associated with overgeneral memory, but not with mood improvement. This is consistent with the possibility that depressed persons may differ from controls in the specificity of personal goals, despite findings suggesting that they may not differ on the number of self-reported approach or avoidance goals [30].

Goals are organised in a hierarchy of increasing specificity from general principles to concrete behaviours, and successful self-regulation requires the individual to formulate specific subgoals and plans that advance progress on more abstract goals [31]. Thus, specific goals are crucial for behavioural self-regulation as they provide more direct links to appropriate action [32], whereas overgeneral goal representations may be more ambiguous. Abstract goals may also be associated with less vivid goal attainment imagery, and this may reduce expectancies of success [33] and the extent to which goals generate anticipatory positive affect [34]. In combination, it is likely therefore that reduced expectancies and anticipatory affect will reduce motivation for goal pursuit.

Given that depression has been characterised in terms of a hypoactive approach system and a hyperactive avoidance system [35], one might hypothesise that reduced goal specificity in depression would emerge for approach but not avoidance goals. On the other hand, overgeneral autobiographical memory in depression obtains in response to both positive and negative retrieval cues [11], whereas another study found specificity deficits across approach and avoidance goals in a mixed anxious-depressed school sample [26]. In this study, we made the more conservative prediction that clinical depression will be associated with the generation of less specific goals across approach and avoidance domains, relative to controls.

In order to be sufficiently motivating, goal representations depend on positive outcome expectancies [31], which may themselves be based upon cognitive appraisals including accessible explanations for goal attainment. Not only do the causal explanations people make for negative events play a crucial role in producing depressive hopelessness [7], but the ease with which a person is able to construct reasons for future events is thought to play a crucial role in the subjective probability of an event [33]. Consistent with this, past research [36] has shown that pessimism about future personal events is related to the proportion of reasons generated for why such events may or may not occur. In another study [37], anxious and depressed individuals gave more (‘pro’) reasons to explain why a negative event would occur than (‘con’) reasons to explain why it would not, the relative number of pro versus con reasons was exactly reversed for positive events, and the relative number of pro and con reasons was associated with likelihood judgements for both kinds of event. A recent study of clinically depressed individuals [30] found no significant difference relative to non-depressed controls in the number of reasons generated either for or against goal attainment. Nevertheless, whereas non-depressed individuals generated significantly more pro than con reasons for goal attainment, depressed individuals did not. However, these studies have not examined specificity differences in individuals' causal explanations for goal outcomes, which may be a crucial determinant in facilitating action toward goal achievement, even when goals are more abstract. Reduced specificity of explanations for, and against, goal success may reflect impoverished representations of pathways toward successful goal achievement. Hence, reduced specificity of explanations for goal attainment may be associated with poorer motivation, commitment and effort towards goal attainment. Our second aim is therefore to investigate whether depressed persons' causal explanations for goal attainment are less specific than those reported by controls.

In summary, we hypothesised that depressed adults would generate less specific goals than never-depressed controls, irrespective of whether their goals are focused on approaching rewarding outcomes or avoiding undesirable outcomes. We also hypothesised that depressed adults would generate less specific (‘pro’) reasons for, and less specific (‘con’) reasons against, goal attainment, relative to controls, across approach and avoidance goal outcomes.