Reynolds Neurocognitive Model of Ethical Decision-Making PDF

Preview

Summary

reading for the course...

Description

Journal of Applied Psychology 2006, Vol. 91, No. 4, 737–748

Copyright 2006 by the American Psychological Association 0021-9010/06/$12.00 DOI: 10.1037/0021-9010.91.4.737

A Neurocognitive Model of the Ethical Decision-Making Process: Implications for Study and Practice Scott J. Reynolds University of Washington Business School The field of business ethics is entrenched in a cognitive approach that portrays the ethical decisionmaking process as a completely deliberate and reasoned exercise. In light of growing concerns about the veracity of this approach, I build upon current knowledge of how the brain functions to present a neurocognitive model of ethical decision making. The model suggests that ethical decision making involves 2 interrelated yet functionally distinct cycles, a reflexive pattern matching cycle and a higher order conscious reasoning cycle, and thereby describes not only reasoned analysis, but also the intuitive and retrospective aspects of ethical decision making. The model sparks research in new areas, holds significant implications for the study of ethical decision making, and provides suggestions for improving ethical behavior in organizations. Keywords: ethical decision making, ethical behavior, neurocognition

Numerous scandals have pushed concerns about individual ethical1 behavior in organizations to the forefront of business activities. As business ethics scholars contribute to this growing dialogue, scholarly models of individual ethical decision making receive more and more attention. These models are not only the foundation for the growing body of empirical research on individual ethical decision making but also serve as a wellspring for prescriptions to organizations (Ford & Richardson, 1994; Low, Ferrell, & Mansfield, 2000; O’Fallon & Butterfield, 2005). Having surveyed the literature, Jones (1991) noted that the majority of the most oft-cited models of ethical decision making reflect the four-stage process originally conceptualized by Rest (1979, 1986). According to Rest, ethical decision making begins with an awareness of the ethical issue. Then the individual makes an ethical judgment, establishes an intention to act ethically, and, finally, engages in ethical behavior. Since Jones’ analysis, other management scholars have presented models, but they too have built upon this same basic four-stage structure (e.g., Gaudine & Thorne, 2001; Thorne & Saunders, 2002). Thus, most ethical decision-making models differ only in their inclusion of other variables considered relevant to this four-stage model. For instance, Hunt and Vitell (1986) argued that social, cultural, economic, and organizational factors shape this process. Similarly, Trevin˜o (1986) argued that factors such as locus of control and the characteristics of the job interact to influence ethical decision making, and Jones argued that characteristics of the ethical issue, its moral intensity, must be considered. While there are certainly advantages of a field converging upon a particular model, convergence can also limit our view (Kuhn,

Scott J. Reynolds, Department of Management and Organization, University of Washington Business School. I thank Sarah Allred, Tara Ceranic, Tom Jones, Terry Mitchell, Linda Trevin˜o, and Jim Weber for their help in the development of this article. Correspondence concerning this article should be addressed to Scott J. Reynolds, University of Washington Business School, Box 353200, Seattle, WA 98195. E-mail: [email protected]

1970). Rest (1986) was very explicit in identifying his model as a cognitive model of ethical decision making—an attempt to understand what individuals think when they encounter ethical issues. Managers, however, often talk about ethical decisions using phrases like “gut-feelings,” “sixth sense” and “intuition” (e.g., Toffler, 1986), which suggests that on some occasions they are not really “thinking” about these issues at all, but are instead acting according to some other mechanism. Recognizing these shortcomings, Haidt (2001) provided compelling arguments challenging the preeminence of the cognitive approach in ethics research, and others have called for the inclusion of such critical ethical decision-making features as instincts, intuition, and retrospection in our models (e.g., Petrinovich, O’Neill, & Jorgensen, 1993; Werhane, 1999; Wilson, 1993). Despite our growing awareness of these problems, we still lack a model that specifies the mechanisms of the ethical decision-making process in such a way that sparks research and directs change in organizations. To guide theory, research, and practice more clearly in this area, I take a neurocognitive approach to ethical decision making. A neurocognitive approach is a research perspective grounded in neuroanatomy, neurophysiology, and neurochemistry (Gazzaniga, 1994, 2000). Neurocognitive research focuses on identifying how the brain literally receives, processes, interprets, and acts upon stimuli. In contrast to a cognitive approach that would focus on identifying what the ethical agent thinks, this neurocognitive model begins by specifying how the ethical agent thinks, literally, from the moment the first stimuli are encountered, through the transmission of electrochemical signals in the brain, to the engagement of ethical behavior. As recent developments in neurocognitive research shed light on the intuitive processes of the brain, this model accounts for both intuitive and deliberate ethical decision making in such a way that it more accurately depicts the ethical decision-making process and, from this extremely grounded position, better directs research and practice into the future. 1

737

In this article I consider the terms ethical and moral to be synonymous.

REYNOLDS

738

This article is divided into four sections. First, I set certain parameters for this discussion. Second, using theory and findings from cognitive neuroscience I present the ethical decision-making process as a two-cycle process of reflexive pattern matching and higher order conscious reasoning. Third, I consider the model’s contributions to ethical decision-making research, namely, its capacity to explain ethical behavior and its implications for the study of ethical decision making. Finally, I conclude by considering the implications of this model for the management of organizations.

Parameters The model requires certain parameters. The first is a definition of an ethical decision. In the same tradition as Jones (1991), I define an ethical decision as a decision that is acceptable to a larger community based on its adherence to moral standards of behavior. As Jones suggested, this definition reflects previous work in crimes and obedience (Kelman & Hamilton, 1989) and is consistent with those definitions that already exist in this literature (e.g., Dubinsky & Loken, 1989; Rest, 1986). While the definition is rather inexact and relativistic, it provides a common frame of reference for analyzing deeper issues. Because this definition has served a large body of research on moral intensity (Frey, 2000), it seems sufficient for this model as well. The second parameter is a common understanding of what represents a neurocognitive approach. Neurocognition, or cognitive neuroscience, is an enormous field of study representing cross-disciplinary work from such fields as physiology, psychology, psychiatry, and linguistics, to name a few (Gazzaniga, 1994, 2000). These areas find common ground in their aim to understand the various operations of the brain, ranging from simple tasks such as vision and recall to the more complex tasks of higher reasoning. As a field of study, cognitive neuroscience employs a wide variety of methodologies, but such methods are generally linked by their theoretical underpinnings. Most neurocognitive researchers utilize neural network modeling—models of brain functioning based on the neuron, the fundamental building block of the brain—to derive working theories of brain functions and to explain experimental results. In this light, the model presented in this article is neurocognitive in its approach in at least two ways. First, it suggests that understanding the function of the neuron is critical to understanding the ethical decision making of the individual, and, second, many of its fundamental arguments are based in the principles of neural network modeling. More specifically, this model relies upon connectionist models of neural network functioning to explain certain aspects of the ethical decision-making process. With these parameters set, we can consider the model in detail.

Two Systems—Two Cycles Scientists have known for decades that different parts of the brain are responsible for different functions and that certain tasks require several different parts of the brain acting jointly (Nolte, 2002). One way to identify different parts of the brain working together on different tasks is to refer to them as systems. Recently, Lieberman, Gaunt, Gilbert, and Trope (2002) distinguished between the X-system and the C-system and posited a model of attribution inference based on the functions and interactions of these two systems. Because these systems are relevant to this

model, I will consider them in some detail, but I refer the reader to their work for a more detailed discussion of these systems.

The X-System—A Reflexive Pattern Matching System The X-system represents those parts of the brain that are most closely associated with nonconscious environmental analysis, what some scholars have described as automatic processing, implicit learning, and even intuition (Haidt, 2001; Lieberman, 2000; Louis & Sutton, 1991; O’Brien-Malone & Maybery, 1998; Taylor & Fiske, 1978). As a complex system, the X-system has many components, but researchers believe that the lateral temporal cortex, amygdala, and basal ganglia are the most relevant to automatic social cognition (Lieberman et al., 2002). In terms of X-system functions, it is presumed that connectionist models of information processing most accurately depict how the X-system manages information (Garson, 2002). Connectionist theory explains the function of the neuron, the fundamental unit of the brain (Churchland, 1989). The neuron is a body cell specialized for the nervous system to carry electrochemical signals. It is capable of identifying a stimulus and transmitting a signal reflective of that stimulus using the natural ionic charges of some of its component proteins. The brain is comprised of perhaps 100 billion neurons, and acting together as a neural network they are able to identify, process, manipulate, and store massive amounts of information. As part of a neural network each individual neuron functions in one of three different types of roles: as input units, as hidden units, or as output units (for a more complete description of connectionist models, see Churchland, 1989). Input units are attuned to very specific stimuli (e.g., visual, auditory). As these neurons encounter stimuli to which they are attuned (e.g., sights and sounds), they are activated and begin a process of transmitting signals reflecting the stimuli to the hidden units. Each neuron communicates the existence of and variation in the stimuli by transmitting different weights, polarities, and strengths of signals to the hidden units. Every hidden unit is connected to multiple input units, and as a hidden unit is activated by various signal weights, polarities, and strengths of numerous input units, it sends a signal on through the system, perhaps to other hidden units. As signals move from neuron to neuron, variations in weights, polarities, and strengths of signals create a signature unique to the form of the stimulus. Eventually, the signals arrive at output units in weights, polarities, and strengths that represent a unique neural pattern associated with the original stimulus. In this sense the system does not process symbolically, but rather subsymbolically, sending just the components or signals of larger symbols across its network. Once the output units have received a neural pattern of the stimulus, that pattern is then assessed against previously formulated base patterns stored in the X-system. Researchers have theorized about the exact nature of these base patterns and have developed several theories about their form (e.g., Medin & Smith, 1984; Smith & Minda, 1998). Generally speaking, scholars agree that these base patterns, which I will refer to in a very generic way as prototypes, can represent sounds, language, objects, scenes, situations, concepts, and even complex social interactions. Nevertheless, current theory (e.g., Prinz, 2002) suggests that prototypes do not necessarily represent these targets. Instead, they can refer to theoretical exemplars of objects, situations, and concepts and thereby create a baseline that is more flexible in its application (for

NEUROCOGNITIVE MODEL OF ETHICAL DECISION MAKING

instance, a “pet” prototype could have qualities of both a dog and a cat). It is important to note that prototypes are multidimensional in form. The brain has five senses at its disposal, all of which can provide input about the nature of any particular stimuli and contribute to the prototype. In addition, prototype dimensions can contain information necessary for action. For instance, the prototype of any object could designate contexts in which the object and its typical role in those contexts are found, and the prototype of a scene could incorporate information about the season in which it occurs or the people who are usually involved (Cantor, Mischel, & Schwartz, 1982; Tversky & Hemenway, 1983). Recent brain research even suggests that our understanding of individual ethical situations can be integrated with the emotional sensations that accompany them (Greene, Sommerville, Nystrom, Darley, & Cohen, 2001). In a sense, the notion of a prototype is consistent with such management concepts as schemas, scripts, plans, categories, implicit theories, and even heuristics (Lord & Maher, 1991). There is one stark difference, though. Management concepts such as these are metaphors crafted to describe decision-making processes. The prototype, however, is not metaphorical—it is literal. It is a pattern of neural electrochemical units that encapsulates the sensory experience to create a material imprint of the external world. It does not describe ethical decision-making methods; it is a mechanism of ethical decision making. Within this system, then, the electrochemical representations of incoming stimuli are matched against prototypes. To the extent that the pattern matches a prototype, the results of the analysis are presented to consciousness. In the event that a match is not made, that result is presented instead. As a reminder, the processes of this system are referred to as “automatic” because they operate reflexively in a nonconscious realm—they are not activated by deliberate thought but instead by the mere presence of stimuli. Amazingly, decision makers are not privy to the inner workings of this system, only to its outcomes. From an evolutionary perspective, such a process is quite cost efficient in the sense that it could be exceptionally debilitating to process information related to all of the stimuli encountered in a given day. Instead, this system operates under the radar of conscious activity, providing only minimal yet necessary information about the environment as prototypes will allow. In everyday experiences, this process allows individuals to recognize elements of the environment (e.g., doors, computer keyboards, receptionists) without expending conscious thought about the nature of those elements. This process is not just about props, though. Ethical situations such as bribery, fraud, misrepresentation, sexual harassment, and the like have prototypical characteristics and are therefore represented in multidimensional structures. When an individual faces a situation, this reflexive pattern matching system is able to conduct a process of nonconscious analysis that accepts stimuli from the environment, transmits and organizes neural patterns reflecting the stimulus, and compares that pattern against prototypes of ethical situations. Thus, when a supervisor, for example, secretly offers a promotion in exchange for sexual favors, the elements of this experience are immediately processed, organized, and matched to an existing prototype of quid pro quo sexual harassment. The situation can then be presented reflexively to consciousness as an ethical (and legal) issue. In this way, ethical prototypes allow decision makers to recognize ethical situations automatically.

739

While ethical prototypes are initially defined by the five senses with information about the observable qualities of the situation, like other prototypes they are multidimensional in form and incorporate added information that provides insights for interaction. Most notably, ethical prototypes include normative evaluations and prescriptive recommendations that help the individual adapt to or cope with the ethical situation. Indeed, it is these additional dimensions that make ethical decisions unique from other kinds of decisions. For instance, a bribery prototype not only describes what a prototypical bribery situation looks like, but also indicates that such situations are viewed by society as ethically abhorrent and that the ethically acceptable behavior is to deny the bribe (Donaldson & Dunfee, 1999). When faced with a situation that resembles a prototypical bribe, the individual not only recognizes the situation in a descriptive sense, but also recognizes society’s general opinion about the matter and has a prescriptive sense of how the situation should be resolved. Thus, the individual not only automatically recognizes that the situation is one of bribery, but also reflexively judges that action as immoral, even though he or she may not be able to explain why (Nisbett & Wilson, 1977). Thus, ethical prototypes are dynamic constructs holding descriptive, evaluative, and prescriptive information in one configuration of neural network signals. In the terms of management research, the processes of the X-system are best conceptualized as a cyclical process of the search for and the structuring of information (see Figure 1). Search is the process of gathering information relevant to a decision (March & Simon, 1958). More literally, it is the process of input neurons being activated by or accepting sensory stimuli from the environment. Structuring, then, is the organization or representation of stimuli in the forms of particular output unit patterns for assessment against prototypes, and it can be a variable or engaging process to the extent that the collective form of the stimuli is ambiguous or vague (Weick, 1995). This is an iterative process in that a search for information or stimuli initiates structuring, but if stimuli are too scarce or too ambiguous for comparative purposes or if the completed pattern is unrecognizable, searching and structuring will continue until the system can determine which, if any, known prototypes match the situation. Then, the reflexive pattern matching system will either present the situation to consciousness as one that matches a stored prototype and is associated with specific normative evaluations and prescriptions, or it will notify consciousness that the situation does not match any known prototypes and further assistance is needed. The process is then handed to consciousness, the C-system, or what I refer to as the higher order conscious reasoning cycle.

The C-System—A Higher Order Conscious Reasoning System While many animals have systems comparable to the X-system, a reflexive pattern matching system, humans are distinct in the extent to which they can process not only perceptions of the environment, but also complex thoughts about those perceptions. The C-system is the mechanism by which such complicated reasoning is accomplished. While researchers know a great deal about the functions of the X-system, the processes of the C-system are so exceptionally complicated that very little is known about how it function...