An Integrative Approach to Consumer Choice

ABSTRACT - Although apparently treated so by the modelers of the two schools, the cognitive-rational and hedonic aspects of choice are not two mutually exclusive elements in consumers' scheme of things. This paper views the two aspects as forming two distinct but consecutive stages of the consumer choice process, differing in terms of their purpose, nature of process, and criteria, as well as relative dominance in a given situation. The discussion concludes with the proposal of an integrative framework.

Citation:

T. C. Srinivasan (1987) ,"An Integrative Approach to Consumer Choice", in NA - Advances in Consumer Research Volume 14, eds. Melanie Wallendorf and Paul Anderson, Provo, UT : Association for Consumer Research, Pages: 96-100.

Advances in Consumer Research Volume 14, 1987      Pages 96-100

AN INTEGRATIVE APPROACH TO CONSUMER CHOICE

T. C. Srinivasan, Vanderbilt University

[The author wishes to thank Terry Elrod for helpful suggestions.]

ABSTRACT -

Although apparently treated so by the modelers of the two schools, the cognitive-rational and hedonic aspects of choice are not two mutually exclusive elements in consumers' scheme of things. This paper views the two aspects as forming two distinct but consecutive stages of the consumer choice process, differing in terms of their purpose, nature of process, and criteria, as well as relative dominance in a given situation. The discussion concludes with the proposal of an integrative framework.

INTRODUCTION

While the study of consumer behavior began to attract the serious attention of scholars as early as 1920's, when the formalization of the marketing discipline and growing interest in advertising research discovered the relevance of consumer behavior, major strides were made during the post-World War II boom in the Americas economy, when the constraint shifted from supply to demand, and the spotlight from the factory onto the marketplace. The study of consumer behavior owes its progress to scholars from a variety of disciplines---economists, decision theorists, cognitive and social psychologists, and phenomenologists. Each group brought the theoretical and methodological orientations of its fundamental discipline to the task, and such chose areas of consumer research most amenable to them. One consequence of this has been that consumer behavior research today has well-developed sub-areas of knowledge which are not always congruent or synergistic.

This paper attempts to look at some aspects of consumer choice process with a view to developing a more basic and integral view of the phenomenon than is prevalent today. The objective is not so much to synthesize the various theoretical approaches, as to evolve a more general, and hopefully better, understanding of consumer choice process.

The paper discusses some of the key constructs employed in the various approaches to consume choice, attempts to integrate some of the more recent findings into a framework, and advances a general model of consumer evaluation and choice process.

CONSUMER BEHAVIOR AS PROBLEM SOLVING

This refers to the currently popular, information processing approach to consumer behavior theory. Here the consumer is viewed as a 'thinker' who goes about solving his consumption problems in a 'rational' and analytical way. The models with this approach have been termed 'cognitive-rational' models because the process is viewed as being i) goal-directed, ii) calculated, and iii) predicated upon some knowledge of costs and benefits of alternative choice (Paul Peter and Tarpey 1975). Central to this paradigm are the concepts of perceived return and perceived risks.

Perceived Return

She concept of perceived return is inherent in the economists' assumption of consumers acting to maximize utility subject to resources constraints. To the rational consumers of the economists' model, utility flowed chiefly through the functional benefits of the performance attributes of a product. The unrealistically restrictive nature of this assumptions was highlighted when Veblen (1899) presented his theory on the social aspects of consumption. The possibility of a third class of benefits, in addition to the functional and social benefits, was postulated when in 1923, Copeland attempted to classify consumer goods on the basis of the rational and/or emotional motives (Horton 1984, p. 7-8).

Today the consumer researchers acknowledge the role of functional, social and personal psychological benefits, although most models do not deal with

Perceived Risk

with the increasing complexity of the economic society, consumption decisions have to be made under greater certainty and there is an ever-increasing demand on people's skills- as consumers- Thus} before they can derive the benefits of consumption of goods, successfully go through a process beset with risk. This risk relates not merely to the uncertainty of the desired outcomes, but the possibility of potentially negative outcomes (Bauer 1967). Further, the uncertainty may not necessarily relate only to the characteristics of the alternatives the consumer faces, but also to his or her buying goals themselves (Moller 1979).

It is obvious that the perceived risk will affect the consumer strategy and that risk reduction has become a major objective of the choice process in the increasing complex consuming environment.

Bounded Rationality

While the concept of perceived risk highlights the 'demand' made on the consumer due to environmental complexity, the concept of bounded rationality focuses on the 'constraint' aspects. Advanced in 1958 (March and Simon 1958), the concept argues that consumers do act rationally when viewed realistically in the light of the constraints.

The first constraint relates to the complexity of the task itself; while a limited amount of complexity is not d but indeed preferred by the it adds to the ' fun' aspect of consumption (this will be referred to later), a high level of complexity is very unsettling and produces an unpleasant psychological state in the consumer (Berkowitz 1969)

The second constraint relates to the ability of the consumer to cope with the complexity of the task. The limitations of the cognitive capabilities of consumers have been dramatically highlighted in the last few decades, when path-breaking technologies have offered the consumers a variety of products which require, for their choice as well as use, advanced cognitive capabilities on the part of the consumer--despite the extension of cognitive capabilities by products like computers (that the very use of computers today requires to perform intricate cognitive gymnastics is another matter!)

The third constraint relates to the profitability of the task; more than ever before, consumers have less discretionary time to spend on consumption decisions and more such decisions to make. Stigler's economics of information theory (Stigler 1961) argues that the consumer will weigh the cost of additional information (in terms of time, money, effort, and delayed consumption) against the value to him, of such information. In fact, McCall (1977) has gone 80 far as to suggest that "convenience may replace price in economic theory".

The value of information in the light of consumers' inclination and cognitive limitations to use it, often seems to be pretty low. In his review of consumer external search, Newman (1977) found little evidence of external search of an extensive nature. The amount of search was very low for an overwhelming proportion of consumers, even for the so-called 'high-involvement' purchases like major appliances.

As a result of such findings, the consumer has increasingly been viewed as trying to satifice rather than maximize utility in consumption decisions.

CONSUMER BEHAVIOR AS HEDONIC CONSUMPTION

This paradigm views the consumer as a 'feeler', who consumes products for certain abstract, intangible, esthetic, symbolic, and hedonic benefits, through a process very different from the one used in the cognitive-rational model.

While sustained and widespread interest in this paradigm is of recent origin, the fact that consumers care for something more than the 'dry' utilitarian, functional benefits has been known for a long time. Copeland's 1923 classification of consumption motives into 'rational' and 'emotional' is recognition of this fact (Horton 1984, p. 8). Among the emotional motives mentioned by Copeland are "Expression of artistic taste," "Happy selection of gifts", Romantic instincts, "Pleasing sense of taste, and -Pleasure of recreation,. In the late forties, and fifties, Dichter and other motivation researchers approached the ultra-utilitarian reasons for consumption from a Freudian, psychoanalytic standpoint (Dichter 1964). In the late fifties and sixties, some valuable contributions were made in the area of product symbolism (Levy 1959, 1963). In the late seventies and eighties, there have been serious efforts to formalize the paradigm, although it is still in the formulative stages; concepts are loosely defined ant/or inconsistently operationalized, research postures are predominantly descriptive, and results have not been organized into a cohesive paradigm" (Hirschman 1981).

The hedonic paradigm differs from the cognitive-rational paradigm along two basic dimensions; it portrays the consumer as looking for benefits qualitatively distinct from the utilitarian, functional ones, through a process fundamentally different from the linear-analytic process of the cognitive-rational paradigm.

Hedonic Benefits

While the utilitarian benefits refer to the functional relationship of the consumer with the product, the hedonic benefits refer to the personal psychological relationship. These benefits range from hedonic (pleasure, fun) and esthetic (beauty) to emotional (happiness, surprise, poignancy) and symbolic (self-identity, self-exploration, self-expression). Ahtola (1985) attempts to distinguish between utilitarian and hedonic aspects of attitude toward a behavior, where "utilitarian aspects...relate to usefulness, value, and wiseness of the behavior as perceived by the consumer. Hedonic aspect relates to pleasure anticipated from the behavior". A more general distinction in terms of functional usefulness vis-a-vis psychological significance of a product, is made by Holbrook (1983) by pointing out "...utilitarian (extrinsic) value tends to result from belief about the way product imagery serves consumption needs whereas esthetic (intrinsic) value teats to hinge on an emotional response to the sign or significant appreciated for its own "sake".

Hedonic Process

The polarization between the two paradigms discussed is at its sharpest in the process aspect. Here we refer to the consumption choice process in addition to the consumption process itself. While the cognitive-rational modelers assume a deliberate, rational, linear, analytical, and essentially cognitive process on the part of the consumer, the hedonic paradigm postulates a spontaneous, emotional, simultaneous, and holistic consumer process which is an integral part of the consumption experience itself. Hirschman (1982) defines hedonic consumption as it refers to " ... consumers" "multisensory images, fantasies and emotional arousal. Hansen (1982), approaches the same issue from findings in the area of brain research, notes that "... most models of consumer behavior imply brain functioning non ally associated with the left brain ... Several misunderstood aspects of consumer behavior can be explained making the use of recent findings on the functioning of the right brain".

In summary, wa growing body of evidence suggests that sensory-emotive stimulation seeking and cognitive information seeking are two independent dimensions", (Hirschman 1982). But, as with cognitive-rational modelers, the hedonic consumption researchers have chosen for investigation, areas of consumption most amenable to their philosophy and methodology, such as the performing arts, the plastic arts, and popular culture products-thus avoiding any overlap with the bailiwick of the cognitive-rational school. This may partly be because the above-mentioned products, exhibiting a high level of 'purity' in terms of their hedonic benefits and process contents, reader their study particularly profitable to an incipient and inchoate paradigm.

CONSUMER CHOICE AS A MULTISTAGE PROCESS

A different aspect of the consumer choice issue relates to the series of mental stages that the consumer is thought to go through in arriving at a purchase decision. Two approaches are found in the literature with respect to this issue.

Hierarchy of Effects

Lavidge and Steiner (1961) proposed that consumers go through a series of stages as shown in Fig. 1. The six steps can be conveniently viewed as being made up of three basic psychological processes: cognitive, affective, and conative (Horton 1984, p. 57).

FIGURE 1

LAVIDGE AND STEINER'S MODEL

Ray (1973) developed this model further, and proposed that consumers use one of three decisions hierarchies based on the above model (differing only in sequence) depending on the buying situation.

Phased Decision Strategy

The cognitive-rational models too, found it necessary to view the consumer choice process as consisting of more than a single stage. Their rationale, however was one of processing efficiency. Lussier and Olshavsky (1979) found that "brand choice strategy is contingent upon task complexity. When three brands were presented, subjects first eliminated unacceptable alternatives and then evaluated the remaining alternatives by a compensatory process".

The use of an eliminative rule such as conjunctive rule in order to reduce task complexity as well as perceived risk and to come up with a set of 'good' alternatives for further consideration has been consistently supported in the -literature (Wright and Barber 1977, Tversky and Kahneman 1981. Grether and Wilde 1984).

AN INTEGRATIVE FRAMEWORK OF CONSUMER CHOICE PROCESS

From the foregoing it can be seen that while cognitive-rational models have concentrated on products.such as detergent, breakfast cereal, TV and washing machine, the hedonic school preferentially study rock concerts, modern dance, historic romance novels, and high fashion products (Mizersky and Mullet 1981, Lapso 1981, Schiffman and Schnarrs 1981). The former group, by ignoring the symbolic aspects of consumption even when they studied products such as automobiles, and the latter, by severely restricting the range of products studied, have tended to give implicit support to the view that the two paradigms are mutually exclusive as they depict the consumer choice process. But it is known that all consumer behavior does contain some symbolic component (Holbrook 1981).

The issue can be -conveniently tackled by using a framework such as the one presented in Fig. 2.

FIGURE 2

ELEMENTS OF CHOICE PROCESS: A FRAMEWORK

The framework distinguishes between the process and outcome aspects of consumer choice. The consumer is hampered in this quest for the process and outcome benefits by the perceived risk and is often daunted and frustrated by the complexity of the process he must go through to make a vise choice. Note that, unlike with outcome, the distinction between the desirable and undesirable aspects (benefits and disbenefits) of the process is not clear cut.

Given such a framework it is easy to see that each situation represents a different equilibrium of these elements. While instances of a predominantly cognitive process have been extensively studied and need no repetition, increasing attention is being paid to buying situations where the emotive aspect dominates. According to Maslow, in some instances, "emotional desires dominate utilitarian motives in the choice of products" (Hirschman 1982). Hansen (1981) notes that "the consumer is not interested in the substantial properties such as ingredients contained in soup or vermicelli. In general, he will rely upon the fact that the soup purchased comes up to a normal standard of quality, and his attitude is influenced only very little by means of product-specific information and the cognitive attitudinal component ... In these areas, the evaluation of the product is more likely to be controlled by emotional appeals". Hirschman (1981) states that the "The symbolic meaning of a product may, in some product classes, overcome or dominate its technical performance as a determinant of consumption ... (in) such diverse products as automobiles, apparel, home furnishings, educational institutions, hairstyle, and leisure time activities". Ahtola (1985) found, la his experiments, that sometimes hedonic aspects dominates ("eating a candy") and sometimes utilitarian aspect dominates ("doing laundry"). Weinberg and Gottwald (1982), investigating impulse purchases, found that "information processing plays a part in the buying decision. But, despite the possibility of later justification of the buying behavior, its influence is smaller than that of emotional engagement".

In this context, it is highly interesting to note that Fishbein's earlier version of the extended model conceptualized three determinants of behavioral intentions, viz. i) attitude toward behavior, ii) social norms, iii) personal norms. Later, he dropped the third factor because it correlated too high with behavioral intention without adding to the predictive power (Fishbein and Ajzen 1975). It is possible that the holistic nature of the process that handles personal psychological variables has rendered- the use of such variables in a multiattribute, compositional model ineffective and somewhat tautologous.

Determinants of Dominance

If it is true that both cognitive-rational and hedonic aspects influence the choice process in most situations, the difference being in their relative dominance, the nest logical question to ask would be what are the factors which determine this dominance. The framework considers three determinants, viz. i) Product variable, ii) Personality variable, iii) Context variable.

Product: This is a rather obvious determinant in that it has, as was previously mentioned, provided the basis for the differing areas of research for the two paradigms. In fact, Hirschman and Holbrook (1982b) classify products into utilitarian and esthetic.

Following the proposed framework, Fig. 3 looks at the product variable along the two dimensions.

FIGURE 3

DETERMINANTS OF DOMINANCE; PRODUCT

Two points are worth noting. The first is the high emotive content and the low cognitive control that mark the impulse good category and the reverse in case of rational goods. In his study of consuming impulses, Rook (1985) notes that "Impulses are action-oriented; they trigger responses quite quickly". The absence of such high emotive content in rational goods is noteworthy. Books on personal selling mention insurance and encyclopedia as two products which should sell themselves on the basis of their rational utility, but are probably the mode difficult products to sell. It is also interesting in this context to see that insurance companies use promotional themes that are high on sentimental and emotional aspects, in order to trigger action.

The other point relates to the complex goods which are high on both utilitarian and hedonic aspects. How the consumers work out their needs along two different dimensions involving different dynamics should be of interest to consumer researchers.

Personality: It is also reasonable to expect some individuals to be more hedonically oriented than others and still others more predisposed toward the cognitive-rational process. For instance, impulse buyers assess themselves as being more emotionalized than others (Weinberg and Gottwald 1982), while some people have a much greater need for cognition than others (Cacioppo and Petty 1982).

More recently, Venkatraman and MacInnes (1985) investigated 'hedonic' and 'cognitive' consumers on their epistemic and sensory exploratory behaviors and on that basis classified consumers as Experience Seekers, Experience Avoiders, Hedonic Individuals, and Cognitive Individuals. Fig. 4 is a schematic adaptation of their findings.

FIGURE 4

DETERMINANTS OF DOMINANCE: PERSONALITY

Context: Apart from product and personality variables, the context of consumption may also influence the relative importance of the cognitive-rational and hedonic dimensions. Imagine a consumer buying a camera i) in his or her capacity as a photographer in the art department of a company, ii) for his or her personal use as a hobbyist and, iii) as a gift for a loved one. The relative dominance of the two dimensions will be very different in each case. Hansen (1981), too, considers this aspect in his study of hemispheral lateralization in relation to consumer behavior. "An important methodological questions is whether relative hemispheral dominance is an individual trait ... or is situational, in that different informational inputs lead to more or less dominating left- or right-brain processes. Hansen (1977) proposes that consumers can make choices in very different ways depending on situational factors and on the nature of the problem" (Hansen 1981).

FIGURE 5

DETERMINANTS OF DOMINANCE: CONTEXT

Fig. 5 attempts to capture the contest effect. Thaler (1985) refers to an example of a person who would not buy an object he feels is attractive but extravagant, but will gladly accept it as a birthday present from his wife, the money in both cases having co e from their joint account. The present framework explains this by the context effect, where the first situation is a cognitive-rational dominant one while gift-giving operates primarily in the affective mode. Thaler (1985) notes the similarity between gift-giving and buyer behavior during vacationing, which, according to this framework, due to the dominance of the hedonic aspects.

A MODEL OF CONSUMER EVALUATION and CHOICE

This section presents a model of consumer evaluation and choice, in the light of the foregoing discussion. The effort is largely speculative and no attempt is made in this paper to test the model.

Central Concepts

The following is a brief discussion of the concepts advanced, which provide the foundation for the model.

Concept of Process Benefits: These refer not only to the 'joy of buying' and the 'fulfillment of the need for cognition', which are experiences sought by the consumer for their own value apart from the outcome benefits of consumption, but also to the disbenefits such as frustration, confusion, feeling of inadequacy, anxiety, fear of loss or disapproval, and torment. The process benefits are important in themselves but even more interesting for their possible impact on the process and thus, the choice made. This distinction has been referred to in the literature variously as buying vs. consuming, choosing vs. using, and purchase vs. product involvement (Holbrook, Chestnut, Oliva, and Greenleaf 1984).

Arousal theory of motivation postulates that an organism seeks to maintain an optimum level of arousal and seeks small, pleasurable ups and towns in the form of stimuli which are moderately uncertain, novel, complex, surprising and changing (Berkowitz 1968). Consumers have described their impulse buying urges as "feeling like 'hungers', 'tingling', 'seem almost physical'.t They use terms such as exciting, risky, a 'surge', fun, naughty, great, happy, exhilarating, satisfying, and compelling" (Rook 1985).

It must be noted that this 'fun' aspect of consumption is severely hampered by perceived risk which generates anxiety. Hence the 'fun' aspect of buying is exemplified most in children, who are largely risk-oblivious. Thus risk reduction is often a precondition for "joy of buying" to operate.

Concept of Reducible Risk: When faced with a perceived risk in a complex consumption environment, the consumer attempts to reduce the perceived risk by processing relevant information, when he feels such 8 step is feasible. But, when he perceives the situation to be too complex or the risk too high to cope with, i.e. when the perceived risk is largely 'non-reducible', he resorts to exogenous methods of choice (such as buying the famous brand or the brand his or her friends have etc.) to reduce the risk. The reducibility of risk would depend on the produce, personality, and context variables.

Concept of Sequential Play of Cognitive and Emotive Processes: Following Lavidge and Steiner's cognitive-affective-conative sequence, it is proposed that the cognitive risk reduction occurs prior to the emotive stage. Following up on the child example ln an earlier paragraph, we can metaphorically say that the 'adult' in the consumer operates first to reduce perceived risk before letting the 'child' in the consumer play ln a relatively risk free environment. Fig. 6 gives a summary of the two stages.

FIGURE 6

CHARACTERISTICS OF COGNITIVE AND EMOTIVE STAGES

Concept of Differing Criteria for the Two Stages: It is important to note that not only the nature of the process involved in the two stages is different but the criteria are different too. One reason is the satisficing approach usually taken by the consumer with respect to risk reduction variables (performance, functional variables).

Secondly, at the end of the first stage, the performance variables have much less diagnostic value having generated a set of alternative that are 'adequate' on these criteria. Dawes defines the conjunctive choice rule with respect to personnel selection as wa selection procedure ... in which an individual is evaluated in his least relevant attribute" (Grether and Wilde 1984). A similar note is struck by Kahneman and Tversky (1979) in their investigation of isolation effect in prospect theory, which refers to "the tendency to cancel out common aspects of choice alternatives and to attend to differences, and can be highlighted by decomposing choice options in sequential manner. For example, in two-stage problems with common outcomes at the first stage, choices are seen to depend on preferences at the second stage conditional to success on the first. That is, since the options are seen to be identical at the first stage, the sequential formulation focusses attention on the second stage. In certain cases, this leads to violations of expected utility theory in that end states are not compared in the normative way (emphasis added).

A third explanation for the differing criteria would be that the processes at play are very different; what are cognitively the most important factors may not be the important ones emotionally.

Concept of Relative Dominance of Processes: It has already been mentioned that the relative dominance of cognitive vis-a-vis hedonic process would depend on i) product, ii) personality and, iii) context. When the perceived risk is too high, an exogenous risk reduction method may be resorted to.

The Model

Nov we are ready to summarize the model in a diagrammatic fashion. Fig. 7 presents the model of consumer choice process given his or her awareness set of alternatives.

FIGURE 7

MODEL OF CONSUMER EVALUATION AND CHOICE

The evoked set is distilled from the awareness set by the application of several criteria such as functional (salience on a major performance variable), social (peer group recommendation), personal (intuitive appeal), or risk reduction (well-known brand). The choice set, consisting of very few brands (often two), is arrived at from the evoked set using a cognitive-rational eliminative process (denoted I) primarily involving performance and social variables. The final choice is made from the choice set by a hedonic process (denoted II), primarily involving personal, psychological variables. The relative dominance of the processes have already been discussed. In industrial buying situations, even the second stage may be a cognitive-rational one, indicated by II-A, where probably a compensatory rule is applied to the choice set brands. On the other hand, in impulse buying situation, the cognitive-rational process may be entirely skipped and the process may involve only the hedonic element ( indicated by I-A). In situations of very high perceived non-reducible risk, the exogenous risk reduction pathway is taken when the consumer goes ahead and chooses the well-known brand, the brand owned by friends, or a personally known auto dealer.

Implications of the Integrative Model

Implications for Research: An explicit concession of stages and processes other than those they are trying to model would enable researchers of either paradigm to explain observed deviations from their models and, in some cases, violation of fundamental theorems (e.g., utility theory). It could also lead to more accurate specification of the phenomenon modeled and even the development of new ways of measurement. Finally, the approach could serve as a more general and, arguably, more realistic framework for looking at some very significant, but apparently immiscible consumer research efforts.

Implications for Marketing Strategy: Two classes of strategic implications are interesting. The first relates to strategic diagnosis. A product may enter the consumer's choice set regularly, but rarely make it as the choice, while another product may fare excellently in being chosen from the choice set, but may get eliminated from the evoked set too frequently. Both these products may exhibit low market shares, but they call for very different corrective action. The second implication relates to the communication of the product benefits. In a buying situation, the consumer may be in his/her risk reduction mode (elimination process) and is likely to be looking closely for the negative points of a brand, while in a non-buying situation (e.g., window shopping) he/she is more likely to be in a relaxed state and to appreciate the positive aspects of a brand. Thus, it may be advantageous for marketers to explore the latter kind of situations for product communication.

CONCLUSION

Although treated so by the modelers of the two schools, the cognitive-rational and hedonic aspects are not two mutually exclusive elements in consumers' scheme of things. This paper is an attempt, in a limited way, to integrate these aspects and present the 'whole picture', as it were.

REFERENCES

Ahtola, Olli T. (1985), Hedonic and Utilitarian Aspects of Consumer Behavior: An Attitudinal Perspective, in Advances in Consumer Research, vol. 12, eds., Elizabeth Hirschman and Morris Holbrook, Provo, UT: Association for Consumer Research, 7-10.

Bauer, Raymond A. (1967), "Some Effect and Persuasibility: A New Look," in Risk Taking and Info ahead and Handling in Consumer Behavior, ed. D. F Cox, Boston, MA: Division of Research, Graduate School of Business Administration, Harvard University.

Berkowitz, Leonard (1969), Social Motivation," in The Handbook of Social Psychology, eds. Gardner Lindzey and Elliot Aronson, Reading, MA: Addison-Wesley.

Cacioppo, John T. and Richard E. Petty (1982), "The Need for Cognition," Journal of Personality and Social Psychology, 42 (January), 116-131.

Dichter, Ernest (1964), Handbook of Consumer Motivations: The Psychology of the World of Objects, New York, NY: McGraw-Hill.

Fishbein, Martin and Icek Ajzen (1975), Belief Attitude, Intention and Behavior, Reading, MA: Addison Wesley.

Grether, David and Louis Wilde (1984), "An Analysis of Conjunctive Choice: Theory and Experiments," Journal of Consumer Research, 10 (March), 373-385.

Remaining references are available upon request from the author.

----------------------------------------