EXTENSIVE ATTENTION: THE EXTENSIVE VS. INTENSIVE ATTENTION MODEL
Translation of the article: Kolańczyk, A. (2011). Uwaga ekstensywna. Model ekstensywności vs. intensywności uwagi. Studia Psychologiczne, 49, 3, 7 – 27.
Abstract
Extensive attention manifests itself in the broadening of the perceptual field. By contrast, intensive attention is narrowly focused and encompasses another perceptual data. Consequently, comprehension of a situation differs these two extreme states of attention. The extensive vs. intensive attention (E-I) model describes: 1) attention filters 2) the range of semantic network activation and 3) motivational factors. Filters are responsible for balancing range and depth of information processing within a limited working memory framework. The presented explanation of E-I attention is largely informed by Posner’s attention network theory and Cowan’s working memory theory. It comprise considerations and research on how paratelic motivation or concentration on general goals determines the broad range of available memory and cognitive receptivity to external stimuli. Extensive attention implies relatively shallow (sensory and schematic) information processing at the cost of deep (semantic) processing. Findings reported by many laboratories suggest that such “broad” attention leads to global information processing. By contrast, intensive attention forms analytical perception. The E-I attention model concerns the way motivational factors, combined with global vs analytical processing, determine the object of attention. As extensive attention has been less explored in psychology, the article is dedicated mostly to this phenomenon.
Introduction
It is easy to confuse extensiveness or breadth of attention with absentmindedness. When we are unable to focus on one thing, we are usually being influenced by other important events and goals and our attention is “flitting” between them and labile. Difficulty concentrating leads, for example, to problems at school. It is important to distinguish between this sort of absentmindedness and extensive attention. Extensive attention is a special state of “lack of concentration” produced by: a) lack of goal (as in relaxation, free exploration or play; Kolańczyk, 1989, 1991) or b) a goal which is general, distant, lacks clarity, a regulation standard which defines desired parameters of functioning (to be a good person, or better than average). Förster, Friedman and Lieberman (2004) found that when we imagine the future we begin to construct representations more abstractly. Therefore, goals which foster extensive attention are typically abstract and, by nature of this abstractness, poorly defined.
Attention processes have long been explained in terms of goal-directed striving (e.g. Allport 1980). They have been thought to help select information which is useful with respect to goals or biological drives. Laboratory research reflects this approach very accurately. The effectiveness of attention selection is usually assessed by means of detection tasks involving the detection of stimuli (e.g. letters, figures or words) in the perceptual field. In other words, goal-determined attitudes (sets) are the key to signal detection. The function of attention processes is to increase the availability of one category (e.g. the letter “A”) at the cost of inhibition of competing categories (other letters). Attention test outcomes depend on the capacity to maintain memory access to specific categories (goal-directed alertness), as in the d2 Test (Brickenkamp 2003) where the sign d” is to be detected.
When, at a given moment, human activity is not being controlled by any specific goal and the setting does not require precise concentration, attention does not disappear, it simply assumes a different state and therefore quality of orientation and object of focus also change. These changes resemble the different states of H2O concentration from ice through water to steam. When ice changes into water we can no longer walk on it and were we to try, we would drown. But water has other qualities. For example, we can swim in it. We cannot test our swimming potential by walking on ice. Similarly, we cannot test the cognitive advantages of extensive states by means of classic detection tests of (intensive/focused) attention although assessment of types of error may offer some important, preliminary suggestions as to their properties. The ideal diagnostic instrument for the assessment of attention must be “fair”. It must enable us to distinguish between extensive and intensive states and take into consideration the cognitive potential of them both.
Meanwhile, all the experimental data on “diffuse” attention have so far been interpreted as detection failure; extensiveness has been analyzed in terms of attention limitation and dysfunction. Lack of concentration on a signal has explicitly been labelled “inattention” (Lau 2011). Even if this is just an operational short-cut (as the author has admitted in personal communication), it is indicative of an entrenched, “local” approach to the study of attention processes. In order to detect a letter, we must be familiar with it and “keep it in mind”. The letter is our object of perception. I also suggest that we anchor our explanation of extensive attention in the object, due to its regulative meaning. What happens to the attention of someone who terminates performance of ongoing tasks and engages in play? First and foremost, his/her meta-motivation changes from goal-directed to paratelic (Apter 1982). In paratelic motivation, action direction largely depends on the situational context. The object of attention is more environmentally determined but it is accessible memory which will determine his/her understanding of the situation.
Broad range attention is clearly an underestimated state although we have more and more evidence supporting its utility. For example, it optimizes the activity of sports men and women engaging in team games (Memmert 2007) and underlies creative intuition (Kolańczyk 1991). In extensive attention the attention filter is permeable and this facilitates the generation of ideas (Mendelson & Griswold 1964; Ansburg & Hill 2003; Howard-Jones & Murray 2003). A person who is focused (e.g. on a game of basketball) may even fail to notice a gorilla in the centre of the visual field (Simons & Chabris 1999) whereas a person who is in a state of extensive attention notices more, albeit differently.
E-I attention means the opposite poles of a dimension which has many intermediary states (Kolańczyk 2009). Changes in the attention parameters lead to qualitative changes in modes of information processing which define at least two extreme “states of concentration”. The rich output of many laboratories allows us to view attention not only as a set of attributes (selectivity, scanning, divisibility) but also as an E-I dimension on which these attributes can change.
The context of attention filter research
If we want to understand “inattentive” extensive attention we must consider how it serves its basic function, which is the selection of information. We may assume that in extensive states the rules of selection are relatively lenient since they do not require much effort (intensive control and inhibition) yet lead to broad perception. Since they take place in conditions of considerable motivational freedom, they help to open “bottom up” processing. Extensive attention does not ruin our comprehension of the world or introduce chaos into the way our minds function, all because the memory structures remain active (we know where we are and with what we are in contact). Chronically aroused memory structures unrelated to sensory impingements – values, incomplete tasks, etc. (to which sensory reality can be assimilated) – play a special role in this process.
In the research on the nature of filtering (Smith & Kosslyn 2009;), theories of late selection which highlight the role of memory in the filtering of meaning may be particularly helpful in explaining extensive attention but this does not mean that selection does not depend, for example, on localisation of stimuli in the visual field (Broadbent 1958), stimulus overload or spatial relations (Khetrapal 2010). Also very informative are studies initiated by Treisman (1960, 1969) who demonstrated the differential availability (activation threshold) of frequently used, contextually associated and important memory contents. Deutsch and Deutsch (1963), meanwhile, developed an inspiring theory of affective attention filter thresholds. According to this theory, stimulus input is affectively “weighed”. We now have evidence of automatic evaluation in the form of differentiation of affective meaning (cf. e.g. Greenwald & Banaji 1995; Ferguson & Bargh 2004) and there is no reason to continue to view the Deutschs’ assumption as illegitimate. Their theory will help me to explain the filter of extensive attention where differentiation of affective arousal is much more subtle than it is in states of high attention concentration (high activation of individual contents at the cost of inhibition of other contents).
Discovery of the flexibility of the attention filter which enables processing to take place at various levels, from sensory processing to signal detection decisions (Johnston 1978), comes closest to the idea I shall propose here. Johnson sought his inspiration in Craik and Lokhart’s theory (1972) of levels of information processing, the theory which also help to explain E-I attention. Johnson argued that people either use shallow (sensory) selection or deep selection (based on semantic operations) depending on the action context. The E-I attention theory (Kolańczyk 1991, 2004b) assumes harmonization of both filters within a stable pool of attention resources. This shared resources assumption is based on the work of both Kahneman (1973) and Szymura and Nęcki (2004). According to E-I attention theory, one filter operates at the cost of the other, i.e. broad sensory processing takes place at the cost of deep semantic processing. Extensive attention involves more sensory and categorical processing and less deep-semantic processing (e.g. in the form of logical reasoning). However, in states of high attention concentration, deep analysis takes place at the cost of sensory processing.
The principles of information selection in Posner’s attention network theory
Posner’s neuropsychological theory of attention (Posner 1980) can help us to understand these selection dynamics and legitimize the E-I model. This theory comprises three relatively independent attention mechanisms or networks: a) the vigilance mechanism and its attention maintenance function, b) the orienting mechanism and its function of (implicitly or explicitly) orienting the senses toward the intake of a particular type of information, and c) the executive mechanism and its function of prioritizing in conditions of cognitive or motivational conflict. This mechanism is “executive” because it subordinates activities to goals, not to their technical realization. The orienting mechanism develops earlier than the executive, goal-directed activity mechanism. It responds to distinctive cues in the entire perceptual field, to their physical properties. When vision is fixated and attention is concentrated in the centre of the visual field, people react to their sensory properties rather than their semantic properties at the peripheries of the visual field (Macrae, Bodenhausen, Milne & Calvini 1999).
The mechanisms responsible for different selective operations are the orienting and the executive mechanisms and the mediators between consecutive filters are disengagement of attention from the object with which it had so far been engaged, its shift to, and finally engagement with a new site (Posner & Cohen 1984).
The three attention mechanisms have been studied from several different perspectives: the neuropsychological, behaviour and functional (cf. Posner & Peterson 1990; Posner & Fan 2007). Even the neurotransmitters and genetic determinants of each of these mechanisms have been investigated (Fan et al. 2001; Fossella et al. 2002). Behaviour research on attention mechanisms has been conducted using standard, signal detection procedures. Participants have been requested to detect objects in central and non-central parts of the visual and auditory fields. Researchers have also compared sensitivity to cues in the perceptual field under conditions of voluntary and involuntary attention control. Arrows or figures priming the target stimulus have been used as cues. According to Posner (1980), voluntary attention and involuntary attention are rooted in different relations between orientation and executive attention. This is important in the context of the present article because extensive states largely involve involuntary attention. Voluntary attention is an endogenous, top-down process. In experimental settings participants themselves decide where to focus their attention and they respond to cues according to accepted principles. Involuntary attention, meanwhile, is exogenous and bottom-up. In experimental settings, participants have to respond to sudden and peripheral signals, even when they do not accurately signal the impending stimulus (Riggio & Kirsner 1997).
In a recent evoked response potential (ERP) study Wang, Wu, Fu and Luo (2010) demonstrated the independence of orienting and attention concentration (the orienting and executive systems) for both voluntary and involuntary attention. This finding is very important for our understanding of E-I attention where there is prevalence of involuntary (E) over voluntary (I) processes. Wang et al. found differences in ERP in response to expected signals, i.e. signals validly primed or not primed by the presence of the target stimulus (75% vs. 50% hits respectively). ERP were slightly different in conditions of subliminal vs. optimal stimulus presentation (which corresponded with involuntary/unconscious vs. voluntary attention). For example, in the voluntary attention setting early orienting responses (N1 and P1) indicated that responses were executive-mechanism-controlled, leading to intensification of perception of anticipated stimuli. We may say that the orienting mechanism was controlled by the executive mechanism. No such changes were observed for involuntary attention, true to the assumptions of Posner’s (1980) theory of attention which posits different relations between orientation and concentration depending on whether attention is voluntary or involuntary. Information selection is determined by the way attention is engaged and hence how orienting and executive mechanisms are cooperating. Various patterns of coordination of orienting and executive selection may therefore be distinguished.
Information filtering in E-I attention states
E-I attention theory (Kolańczyk 1991, 2004b, 2009a) draws from the research findings presented above and the conclusion that increased executive mechanism activity means increased control of this mechanism by the orienting mechanism. In this case, information is processed deeply and analytically, blocking out other information, both sensory and semantic. Intensive attention is shaped by task-oriented motivation. Together with changes in meta-motivational states toward paratelic motivational states, the attention orienting mechanism gains regulative autonomy and the processing of information input from the perceptual field broadens at the cost of deep semantic analysis. Cognitive processes are now more sensory and schematic. Thanks to the operation of orienting network filters which are poorly restricted by the executive mechanism, extensive attention has such a broad range. Intensive attention is determined by the executive network filter which controls the execution of tasks from the semantic level. Hence extensive attention is broad and shallow and information processing in this state is sensory and schematic whereas intensive attention is narrow and information processing in this state is deep and semantic.