Sample Cognitive Control Research Paper. Browse other research paper examples and check the list of research paper topics for more inspiration. iResearchNet offers academic assignment help for students all over the world: writing from scratch, editing, proofreading, problem solving, from essays to dissertations, from humanities to STEM. We offer full confidentiality, safe payment, originality, and money-back guarantee. Secure your academic success with our risk-free services.
Control refers to the ability to direct mental function and behavior in accord with an internally represented set of intentions. This is manifest in higher cognitive function in many forms: the ability to direct attention to a specific stimulus within a large array of other competing, and perhaps more salient stimuli (e.g., finding the face you are looking for within a crowd), maintaining a new and important piece of information in mind against distraction (e.g., remembering a telephone number you just got from directory assistance until you dial, while in a phone booth at a noisy airport terminal), overcoming a compelling but undesirable behavior (e.g., suppressing the urge to scratch a mosquito bite) or pursuing a complex but unfamiliar behavior (e.g., playing a new piano piece) and, perhaps most importantly, the ability to respond flexibly and productively in novel circumstances (e.g., mentally explore the consequences of a complex sequence of moves in a game of chess).
Academic Writing, Editing, Proofreading, And Problem Solving Services
Get 10% OFF with 24START discount code
The distinction between controlled and automatic processing is one of the central concepts within modern cognitive psychology. Controlled processing is considered to be effortful, and to rely on a limited capacity system, while automatic processing is assumed to occur independently of this system (Posner and Snyder 1975, Shiffrin and Schneider 1977). This concurs with common experiences, such as the ability to carry on a conversation while driving a car (an automatic process) but not while conducting multi-digit arithmetic in one’s head (a process that relies on control). Many contemporary theories posit that there is actually a continuum between controlled and automatic processing (for a review, see Cohen et al. 1990). Nevertheless, virtually all theorists acknowledge the need for some mechanism, or set of mechanisms, responsible for the coordination of processing in a flexible fashion—particularly in novel or demanding tasks. This idea figures centrally in Baddeley’s classic theory of working memory (Baddeley 1986), which postulates two critical components: a storage component responsible for the active maintenance of information in a short-term store, and a executive control component responsible for the manipulation and coordinated use of this information. For example, in a multi-digit multiplication problem, the storage component maintains the intermediate products while the executive carries out the arithmetic operations.
The postulation of a central executive closely paralleled theorizing regarding the nature of frontal lobe function (for a review, see Shallice 1982), based on the clinical observation that patients with frontal lesions exhibit a ‘dysexecutive syndrome.’ The frontal lobes are the area of the brain most highly expanded in humans relative to other species (see Fig. 1). Damage to this area is associated with impairments in characteristically human cognitive functions that are directly dependent upon control (such as planning, or the ability to respond adaptively in novel circumstances), and disturbances of this structure have been implicated consistently in neuropsychiatric diseases such as schizophrenia that also appear to be uniquely human. Indeed, the earliest neurologists, neuroscientists and neuropsychologists recognized the importance of the frontal lobes in the control of behavior. Perhaps this observation was made most dramatically in the classic case of Phineas Gage, the unfortunate railroad foreman who suffered damage to his prefrontal cortex (PFC) when a 1-1 2 -diameter rod penetrated his skull in a construction accident. Originally someone who was considered to be thoughtful, responsible, and of sound judgment, following the accident he was de-scribed as ‘capricious … and unable to settle on any of the plans he devised for future action’ (Harlow 1848). Such changes have been observed repeatedly in patients with damage to the frontal cortex (for a review, see Stuss and Benson 1986). Furthermore, patients with frontal damage perform poorly in tasks that require even the simplest forms of cognitive control, such as the Wisconsin Card Sort Task (WCST) (Milner 1964), and, most recently, brain imaging studies consistently have revealed increased activity of the prefrontal cortex while subjects are performing tasks that demand cognitive control (for a review, see Miller and Cohen 2001). Although this accumulation of findings has led most investigators to assume that the prefrontal cortex plays a critical role in cognitive control, they have provided little insight into the specific contribution that it makes, or the mechanisms by which it operates. However, more detailed neurobiological studies have begun to shed light on this question.
Using single unit recording techniques, Fuster and Alexander (1971) and Niki (1974) reported the remarkable finding that some neurons in the prefrontal cortex continue to fire during the delay between a stimulus and a contingent response, even after the stimulus has disappeared. Goldman-Rakic and her colleagues followed up on this finding and, in a series of elegant studies, demonstrated that the firing of such neurons was stimulus-selective, and was critical for performance in delayed-response tasks (Goldman-Rakic 1990).
These findings strongly implicate neuronal activity in PFC as the site of temporary storage of information in working memory. Recent neuroimaging studies of humans have provided strong convergent support for this idea, demonstrating sustained activity in PFC during the performance of simple working memory tasks (for a review, see Smith and Jonides 1999). However, this discovery seemed to pose an interesting puzzle. Most of the neuropsychological data suggested that the PFC housed the central executive which, at least within Baddeley’s influential theory of working memory, was clearly distinguished from the storage component thought by most to be housed primarily in more posterior structures. Computational models of cognitive control have offered a different perspective on this problem that may reconcile the role of the PFC in control and storage.
1. Computational Models
Shallice (1982) presented the first account of the role of PFC in cognitive control within an explicitly computational framework. He proposed that the PFC housed a supervisory attentional system (SAS)—a mechanism by which PFC coordinates complex cognitive processes. Shallice’s theory was described in terms of a production system architecture. This has appeal, as it relates the executive functions of frontal cortex to the well characterized mechanisms of other production system theories, which include the active representation of goal states to coordinate the sequences of production firings involved in complex behaviors (e.g., Anderson 1983). One feature of goal representations is that they must be maintained actively throughout the course of a sequence of actions to direct behavior effectively. This coincides with the observation that PFC appears to be specialized for the active maintenance of task-relevant information. Thus, it is possible that PFC is specialized for the maintenance of a particular type of information, as a means of executing control over performance.
While Shallice’s theory was not implemented originally as a functioning model, Kimberg and Farah (1993) proposed a model of frontal function—also using a production system architecture—that simulated performance in a variety of tasks considered to rely on frontal lobe function (including the WCST), and illustrated that damage to this component of the model produced impairments similar to those observed in patients with frontal damage. However, while such models have produced insights into the functional role of PFC in cognitive control, their components do not have a transparent mapping on to specific neurobiological mechanisms. Neurobiological plausibility is not a requirement, per se, of a theory that seeks to explain the cognitive functions of PFC. Nevertheless, the question of how goal directed behavior arises from the firing of millions of neurons presents a mystery in its own right. Furthermore, under-standing how different forms of cognitive impairment arise from different forms of damage to PFC (e.g., the effects of stroke or injury vs. schizophrenia) provides an important motivation for understanding how cognitive control may arise from the specific neurobiological mechanisms housed within PFC.
Recently, investigators have begun to use neural network models to better understand how PFC may carry out its functions. Such models (also known as connectionist, or parallel distributing processing models) simulate the behavioral performance of human subjects (or animals) in cognitive tasks using neutrally-plausible processing mechanisms (e.g., the spread of activity among simple processing units along weighted connections; see Rumelhart and McClelland 1986). The goal of this effort is to identify principles of neural function that are most relevant to behavior. Using this approach, Dehaene and Changeux (1989), Levine and Prueitt (1989), and Cohen and Servan-Schreiber (1992) have all described models of pre-frontal function, and used these to simulate the performance of normal and frontally-damaged patients in tasks that are sensitive to PFC damage, such as the WCST and others. All of these models simulate PFC function as the activation of a set of units that represent the ‘rules’ of the task; that is, units whose activation leads to a task-relevant response, even when this may not be the one most strongly associated with the stimulus. However, in most models, the PFC units themselves are not responsible for generating the response directly. Rather, they influence the activity of other units whose responsibility this is. This is clearly illustrated by a model of the Stroop task developed by Cohen et al. (1990).
In the Stroop task (Stroop 1935), subjects respond to words either by reading them, or by naming the color in which they are displayed. In the critical condition, the word itself conflicts with the color in which it is displayed (e.g., the word GREEN displayed in red). Subjects have no trouble reading such words (e.g., saying ‘green’). However, when they are asked to name the color (e.g., say ‘red’), they are significantly slower, and sometimes even make errors. This reflects the highly practiced, and therefore prepotent tendency to read written words, which interferes with the ability to name the color of a Stroop conflict stimulus. The ability to name the color in the face of such interference (that is, to produce the weaker, but task-relevant response) is a simple but clear example of cognitive control.
Cohen et al. (1990) constructed a model of this task, as shown in Fig. 2. The connections among the units in this model defined two processing pathways, one for word reading and another for color naming. The connections in the word-reading pathway were stronger, capturing the assumption that this was the more practiced task. Because of these stronger connections, information flowing along the word pathway interfered with color naming, simulating the interference effects observed when human subjects perform this task. Indeed, the model’s ability to produce a response to the color in the face of such interference required the addition of a set of units (labeled ‘Dimension’ in Fig. 2), which provided additional activation of the units in the color-naming pathway. This additional activation biased processing in favor of this pathway, allowing it to compete more effectively with, and prevail over, activity flowing along the stronger word pathway. This biasing effect corresponds precisely to the role of top-down control in neurophysiological models of attention (such as the Biased Competition Model of Desimone and Duncan 1995), and has been proposed as a mechanism by which PFC exerts control over processing (Cohen and Servan-Schreiber 1992, Miller and Cohen 2001). Models such as this have been used to simulate the performance of both normal subjects and patients with frontal damage in a wide range of tasks that tap cognitive functions commonly associated with PFC function, such as working memory, attention, behavioral inhibition, planning, and problem solving.
2. Guided Activation As A Method Of Control
The Stroop model brings several features of PFC function into focus. First, it emphasizes the view that the role of the PFC in control is modulatory, guiding the flow of activity along pathways in other parts of the brain that are responsible for task performance. For example, activating the color unit does not in itself transmit information about a particular response (red or green). Rather, it simply insures that activity flowing along the color-naming pathway will have a greater influence over the response than activity flowing along the word pathway. In this way, representations within the PFC can function as intentions, rules, or goals (comparable to those in a production system architecture), by setting up the appropriate relationship between a stimulus (or category of stimuli) and an associated response (or set of responses) through the proper guidance of activity along path-ways in other parts of the brain. Recent neurophysiological findings regarding the firing properties of neurons in PFC provide strong support for this view (for a review, see Miller and Cohen 2001). Recent neuroimaging studies also provide convergent sup-port, indicating that PFC activity occurs when behavior relies upon explicit knowledge about rules or arbitrarily determined conjunctions of stimulus features (for a review, see Miller and Cohen 2001). Note that this function is not necessarily restricted to mappings from stimuli to responses, but applies equally well to mappings involving internal states (e.g., thoughts, memories, emotions, etc.). Note also that this function is not necessarily unique to PFC. There may also be more local forms of control, responsive to regionally specific needs for the biasing of processing. However, the wide range of anatomic connections that the PFC shares with virtually all associative areas of the brain places it in a strategic position to guide the flow of activity between and among these other regions (for a review, see Miller and Cohen 2001)—a position well suited to its presumed role in the control of higher cognitive processes.
The emphasis of this guided activation model of PFC on its modulation of other brain areas actually responsible for task execution is consistent with the classic pattern of neuropsychological deficits associated with frontal lobe damage: The individual elements of a complex behavior are usually left intact, but the subject is not able to coordinate them in a task-appropriate way (for example, a patient who, when preparing coffee, first stirred and then added cream; Shallice 1982). The guided activation model also captures another critical feature of theories about the role of PFC in executive function: the importance of sustained activity as a critical component of control. For a representation to have a biasing influence, it must be activated. For this influence to endure (e.g., over the course of performing a task), it’s activity must be sustained. For example, to continue color naming, the activity of the color unit must be maintained, lest the word begin to dominate processing. Similarly, an increase in the demand for control requires greater or more enduring activity of the corresponding units in PFC. This concurs with accumulating evidence from the neuroimaging literature, that tasks thought to rely on controlled processing consistently engage the PFC (for a review, see Smith and Jonides 1999). It is also consistent with both behavioral and neuroimaging findings regarding the effects of practice on automaticity and the involvement of PFC. Increased practice on a task should strengthen its underlying pathway, reducing its reliance on control. Indeed, consistent practice on a task reduces the amount of PFC activity observed, and PFC damage can impair new learning but spare performance on well-practiced tasks (for a review, see Miller and Cohen 2001).
Finally, this approach helps to unify the role that PFC plays in the variety of cognitive functions with which it has been associated, such as selective attention, behavioral inhibition, and executive function in working memory. These can all be seen as varying reflections, in behavior, of the operation of a single underlying mechanism of cognitive control: the biasing effects of representations in PFC on processing in pathways responsible for task performance. For ex-ample, selective attention and behavioral inhibition may be viewed as two sides of the same coin: Attention is the effect of biasing competition in favor of task-relevant information, and inhibition is the consequence that this has for the irrelevant information (cf. the Biased Competition Model of Desimone and Duncan 1995). This assumes that inhibition occurs due to local competition between conflicting representations (e.g., between the two responses in the Stroop model), rather than centrally by the PFC. The ‘binding’ function of selective attention can also be explained by such a mechanism, by assuming that PFC representations can select the desired combination of stimulus features (over other competing combinations) to be mapped on to a response.
3. Outstanding Questions
The guided activation model provides an integrative, and mechanistically explicit framework for considering the role of PFC in cognitive control. At the same time, it brings into focus a number of important and unanswered questions. For example, despite the longstanding observation of delay-period activity for PFC neurons, and the centrality of this property in neural network models of PFC, little is known about the actual mechanisms by which neuronal activity is sustained. This could reflect a cellular property of PFC neurons (e.g., bistability), or a circuit-level phenomenon (e.g., recirculation of activity within the PFC, or between PFC and other structures). Several models have proposed the latter, assuming that representations are maintained in PFC as attractor states (for a review, see O’Reilly et al. 1999). However, neurobiological studies are needed to confirm this hypothesis. A closely related question concerns the mechanisms by which patterns of activity are updated in PFC. These must be able to satisfy two conflicting demands: On the one hand, they must be responsive to relevant changes in the environment; and on the other, they must be resistant to updating by irrelevant changes. Neurophysiological studies suggest that PFC representations are selectively responsive to task-relevant stimuli and robust to interference from distractors (for a review, see Miller and Cohen 2001). Not surprisingly, two hallmarks of damage to PFC are perseveration (failure to update) and distractibility (inappropriate updating). These observations suggest the operation of mechanisms that insure the appropriate updating of PFC activity in response to behavioral demands.
Recent modeling work has suggested that brainstem dopaminergic systems and the basal ganglia may play an important role both in updating representations in PFC and learning how and when to do so (for a review, see Miller and Cohen 2001). Other work has suggested that the anterior cingulate cortex—a midline structure within the frontal lobes—may play an important role in monitoring task performance, and identifying the need to allocate control (Botvinick et al. 2001). Such theories make important predictions about the neural mechanisms underlying cognitive control that serve as valuable challenges to future neurobiological research in this area.
An equally fundamental question concerns the nature of representations in PFC and how they arise. For example, the Stroop model assumes there are units that represent each of the two dimensions of the stimulus (color and word), with the appropriate connections to the corresponding pathways. However, are there such units in the PFC for every possible combination of stimulus and response of which tasks may be composed? This seems unlikely. Yet, there must be sufficient representational richness to support the flexibility in behavior that the PFC seems to afford. What principles define this set of representations, how they are organized, and how they are learned are important questions at both the computational and neurobiological levels.
A better understanding of the mechanisms under-lying active maintenance may also provide insight into one of the most striking and perplexing properties of cognitive control: its severely limited capacity. This has long been recognized in cognitive psychology (Posner and Snyder 1975, Shiffrin and Schneider 1977), and is painfully apparent to anyone who has tried to talk on the phone and read email at the same time. The resource limitation of cognitive control has played an explanatory role in many important models of human cognition. However, to date, no theory has provided an explanation of the limitation itself. This is a sine qua non of cognitive control, and therefore provides an important benchmark for any theory that seeks to explain its underlying mechanisms.
Finally, it is important to recognize that the PFC is certainly not the only brain structure involved in cognitive control. As noted above, mechanisms similar to those within PFC may operate locally in other parts of the brain. Furthermore, there are certainly other types of mechanism critical to cognitive control. For example, the mechanisms responsible for keeping an immediate goal in mind (e.g., working on a book chapter) are not likely to be the same ones responsible for realizing long-term goals (e.g., getting the book published). While the former may be guided by representations actively maintained in PFC, the latter almost certainly engage mechanisms of long-term storage. Some suggestions have been made about how the PFC may interact with the hippocampus to orchestrate the storage and retrieval of goal representations at appropriate times (O’Reilly et al. 1999), however, this remains another area in need of further research.
4. Conclusion
One of the great mysteries of the brain is how purposeful, goal-directed behavior emerges from the millions of relatively simple processing units that are its basic computational elements. Behavioral, neuropsychological, and neurobiological data converge on the idea that the frontal lobes play a critical role in cognitive control. Neural network models have begun to suggest how this may be carried out, as sustained patterns of activity within PFC modulate, or ‘guide’ the flow of activity along pathways in other parts of the brain responsible for performing a task. However, many fundamental questions remain to be addressed. The human brain is arguably the most complex device in the known universe, and its capacity for higher cognitive function continues to be one of its deepest mysteries. Unraveling this mystery stands as one of the most exciting challenges in science, and the rapid development of sophisticated new empirical methods (such as functional brain imaging) and theoretical tools (such as neural network modeling) offer hope that this challenge can be met.
Bibliography:
- Anderson J R 1983 The Architecture of Cognition. Harvard University Press, Cambridge, MA
- Baddeley A 1986 Working Memory. Clarendon Press, Oxford, UK
- Botvinick M M, Braver T S, Carter C S, Barch D M, Cohen J D 2001 Conflict monitoring and cognitive control. Psychological Review
- Cohen J D, Dunbar K, McClelland J L 1990 On the control of automatic processes: A parallel distributed processing account of the Stroop eff Psychology Review 97: 332–61
- Cohen J D, Servan-Schreiber D 1992 Context, cortex and dopamine: A connectionist approach to behavior and biology in schizophrenia. Psychology Review 99: 45–77
- Dehaene S, Changeux J P 1989 A simple model of prefrontal cortex function in delayed-response tasks. Journal of Cognitive Neuroscience 1: 244–61
- Desimone R, Duncan J 1995 Neural mechanisms of selective visual attention. Annual Review of Neuroscience 18: 193–222
- Fuster J M, Alexander G E 1971 Neuron activity related to short-term memory. Science 173: 652–4
- Goldman-Rakic P S 1990 Cellular and circuit basis of working memory in prefrontal cortex of nonhuman primates. Progress in Brain Research 85: 325–35
- Harlow J M 1848 Passage of an iron rod through the head. Boston Medical and Surgical Journal 39: 389–93
- Kimberg D Y, Farah M J 1993 A unified account of cognitive impairments following frontal lobe damage: The role of working memory in complex organized behavior. Journal of Experimental Psychology 122: 411–28
- Levine D S, Prueitt P S 1989 Modeling some effects of frontal lobe damage—novelty and perseveration. Neural Networks 2: 103–16
- Miller E K, Cohen J D 2001 An integrative theory of prefrontal cortex function. Annual Review of Neuroscience 24: 167–202
- Milner B 1964 Some effects of frontal lobectomy in man. In: Warren J M, Akert K (eds.) The Frontal Granual Cortex and Behavior. McGraw-Hill, New York, pp. 313–31
- Niki H 1974 Prefrontal unit activity during delayed alternation in the monkey. 1. Relation to direction of response. Brain Research 68: 185–96
- O’Reilly R C, Braver T S, Cohen J D 1999 A biologically-based computational model of working memory. In: Miyake A, Shah P (eds.) Models of Working Memory: Mechanisms of Active Maintenance and Executive Control. Cambridge Uni-versity Press, New York
- Posner M I, Snyder C R R 1975 Attention and cognitive control. In: Solso R L (ed.) Information Processing and Cognition. Erlbaum, Hillsdale, NJ
- Rumelhart D E, McClelland J L 1986 Parallel Distributed Processing: Explorations in the Microstructure of Cognition. MIT Press, Cambridge, MA
- Shallice T 1982 Specific impairments of planning. Philos. Trans. R. Soc. London Ser. B 298: 199–209
- Shiffrin R M, Schneider W 1977 Controlled and automatic human information processing: II. Perceptual learning automaticity, attending and a general theory. Psychology Review 84: 127–90
- Smith E E, Jonides J 1999 Storage and executive processes in the frontal lobes. Science 283: 1657–61
- Stroop J R 1935 Studies of interference in serial verbal reactions. Journal of Experimental Psychology 18: 643–62
- Stuss D T, Benson D F 1986 The Frontal Lobes. Raven, New York