I consider the title of this section of the conference, “Theoretical Reasons for a Choice” (between the laboratory and everyday life), to be a bit misleading because, really, there is no necessity for choice. The issue should be considered from the perspective of adopting sampling strategies that will support the scientific generalization to be drawn: Do the sampling procedures support the inference from the particular samples chosen to the universes of interest?

Brunswikian representative design

To begin, I shall briefly describe some Brunswikian views regarding research strategies and develop these ideas within a sampling framework. Brunswik (1952, 1956) argued that we should build a science adequate to understand the behavior of organisms in their environment and should adopt research strategies appropriate to that undertaking. Because behavior takes place in a semichaotic medium that contains cues of limited trustworthiness, expressed vicariously, a research strategy different from those usually advocated will be necessary to realize this understanding. This strategy is called probabilistic functionalism; it utilizes representative design and is adequate to the task of conceptualizing and understanding complex behavior (Petrinovich, 1979). Unfortunately, it is not a simple conceptualization, nor is it easy to implement.

The Brunswikian argument can be conceived as an exercise in sampling theory. First, there is agreement between proponents of representative design and systematic design that it is essential to obtain a representative sample of subjects on which to base theoretical conclusions of general applicability.

Comprehension of spoken language involves rapid construction of meaning from a transitory acoustic signal, the complexity of which can easily be overlooked. In ordinary conversation, speech rates may average between 100 and 180 words per minute (wpm), and a speaker reading aloud can easily average over 200 wpm. In addition, the words in spoken discourse often are unclear or garbled (Pollack & Pickett, 1963). In spite of these challenges, phonemic and syntactic structures interact with semantic and contextual contraints to produce the perception of an intelligible message in “real time” (Marslen-Wilson & Tyler, 1980). That is, unlike reading, in which the viewer may backtrack and proceed at a comfortable rate, speech is heard at the rate produced by the speaker. Not only must this complex acoustic signal be understood phonologically at this extraordinary rate, but also the utterances must be further analyzed into the sentences and propositional representations that give rise to meaning.

Whereas older adults often suffer from deficits in auditory processing (Olsho, Harkins, & Lenhardt, 1985), it has also been argued that they have particular difficulty with tasks requiring “deeper,” more effortful processing operations (Craik & Simon, 1980) and are slower in performing many cognitive operations (Salthouse, 1980, 1982). For these reasons, it might seem surprising that older adults are not more often noticed to have trouble in understanding everyday speech (e.g., in conversation, from television or radio).

Management techniques used in hospitals will vary according to the type of hospital ward and the clinical diagnoses of the patients. Different programs are used on acute general medical wards, rehabilitation wards, and long-stay wards. Different approaches are taken according to whether an impairment is one that is likely, to some extent, to show improvement (stroke or early head injury), to remain static (longer-term head injury), or to deteriorate (multiple sclerosis and dementia). Most of the previous literature has reported studies based on rehabilitation or geriatric wards. These will be considered in this chapter, but most emphasis will be placed on approaches that might be used on general medical wards, as there have been few published reports on the training of memory in this setting.

The studies considered are those that were designed for real-life memory problems, as opposed to investigations of the ability to learn experimental material. The former studies fall into two main categories. On geriatric or long-stay wards, the main approaches used are reality orientation, environmental modification, and reminiscence therapy. In a rehabilitation setting, training in specific memory strategies, such as visual imagery (Glasgow, Zeiss, Barrera, & Lewinsohn, 1977; Wilson 1981, 1982), PQRST (Glasgow et al., 1977), and first-letter cuing (Wilson, 1982), is more often used. These strategies may be carried out with individuals or in groups (Wilson & Moffat, 1984a).

In the course of this chapter, I hope to convince the reader that research methods in cognitive psychology have a great deal to gain by adopting a view of the world based on general systems theory (GST). The study of memory should be no exception. I hope to describe what GST is, how it evolved, and what some core themes and systems functions might be.

The human is a system, I shall argue, and it is nested in social and environmental systems. Memory serves as a control for the availability and flow of information and energy with the human system and between systems, balancing and regulating stimulation. So, I shall argue, memory should be studied keeping in mind its overall purpose, that is, control of the overall daily inner and outer environments in a world in which the present copies the past to some degree.

For memory research, this implies the inclusion of several levels of everyday tasks and the broadening of research questions to include several nested systems. I hope to demonstrate through examples that research questions related to aging make more sense when framed in terms of GST ideas, such as “control of stimulation,” “compensatory strategies,” “boundary flexibility,” and “entropy,” because the systems approach is larger than most other models or approaches. I shall be emphasizing David Rubin's “why” question, although “how” and “when” will be touched on.

In Part III, we consider ways of compensating for everyday memory failures and discuss methods for improving memory. We begin by looking at some of the issues involved, continue with reports of studies designed to enhance memory in the normal elderly, and conclude by examining memory programs for those with acquired brain damage.

The opening chapter in Part III is by Lars Bäckman, of the University of Umea, Sweden. He discusses types of compensation strategies used by older adults in episodic remembering. Bäckman provides a scholarly text that suggests that although the elderly do less well on many memory tasks than younger people, they are nevertheless able to compensate for some of their memory deficits. They do this in three major ways. First, they may pick up on the support provided by experimenters. For example, scores on paired-associate learning tasks can be improved by responding positively to a tester 's prompts and cues. Second, compensation may occur through properties inherent in the task itself. Thus, for example, when elderly subjects are asked to recall a series of actions they have watched, their recall is inferior to that of younger subjects, but if the elderly subjects perform the actions themselves, the age effect is eliminated. Third, the elderly may compensate through cognitive support systems. It can be argued, for instance, that elderly chess players compensate for encoding and retrieval deficits by better global evaluation of positions.

Although the vast majority of experiments on latent inhibition have used animals as subjects, there has been increasing interest in the phenomenon as it occurs in humans. Because the normal procedures for producing latent inhibition are effective with children, but not with adults, who require a masking task, it is convenient to examine separately these two populations of subjects.

Children

The study of stimulus preexposure effects, using children as subjects, has not received much attention in the literature. Indeed, the work can be assigned almost in toto to the activities of three different laboratories. The earliest research, that of Cantor and his associates, concentrated on the effects of stimulus familiarity on reaction time (Cantor, 1969a,b). These efforts were continued by Kraut and Smothergill (e.g., Kraut, 1976; Kraut & Smothergill, 1980; Smothergill & Kraut, 1980), who started from a cognitive theoretical base (Posner, 1978; Posner & Boies, 1971). On the other hand, Lubow and his colleagues (e.g., Lubow, Alek, & Arzy, 1975; Kaniel & Lubow, 1986) have explored stimulus familiarity primarily within the context of its effects on subsequent learning. Because there is general agreement that the stimulus preexposure effect that is termed “latent inhibition” involves a learning deficit, the reaction-time studies will be omitted from this review.

Whereas the stimulus familiarization effect is based on reaction-time studies, stimulus preexposure also has decremental effects on later learning when the previously familiarized stimulus is employed as a conditioned or a discriminative stimulus, one to which new associations must be acquired.

How can one explain the latent inhibition phenomenon? Why does nonreinforced stimulus preexposure of the to-be-conditioned stimulus result in a decrement in associability to that stimulus as compared with another stimulus that has not been preexposed? Thus far, this book has provided a description of the means whereby one can produce latent inhibition, attenuate latent inhibition, and even obliterate latent inhibition. If these conditions can be divided into those that are necessary and those that are sufficient, perhaps that is explanation enough. However, researchers have not always allowed themselves the comfort of such readily attained descriptions, but have sought other explanations for this phenomenon, usually either in neurophysiology, hypothetical (Hebb, 1955) or real, or in behavior, where “explanation” means “related to other behavioral concepts and/or empirical laws.” Thus, for conceptual nervous system type explanations, writers interested in latent inhibition have appealed to habituation, particularly of the orienting response (Maltzman & Raskin, 1965; Wolff & Maltzman, 1968), and filter-type attention mechanisms (Ackil et al., 1969; Carlton & Vogel, 1967; Siegel, 1969a). More recently, as we saw in the last chapter, there has been a considerable amount of work on the real nervous system in regard to latent inhibition, with accompanying theoretical considerations.

Those inclined to find explanations within behavior theory have inspected the possibilities of conditioned inhibition (Reiss & Wagner, 1972; Rescorla, 1971) and competing or complementary responses (Lubow & Moore, 1959; Lubow et al., 1968).

Modern science has developed to such a point that there are myriad research areas, dark corners as well as bright little chambers, that are penetrated only by a knowledgeable few who happen to be working in a given field – the specialists. I would not be surprised, then, to discover that most of my colleagues, whether they label themselves psychologists, psychobiologists, behavioral scientists, or neuroscientists, do not know what “latent inhibition” is, or at best confuse the term with older research areas that are indeed its distant cousins – latent learning and conditioned inhibition. It is for these readers that I begin this treatise with a definition, as simple as it may be.

“Latent inhibition” is defined by three characteristics. One is concerned with conditions for producing it, the second with the conditions for measuring the effect, and the third with the direction of the differences between groups. More specifically, latent inhibition is the detrimental effect of passive, nonreinforced preexposure of a stimulus on the subsequent ability of an organism to form new associations to that stimulus. To demonstrate latent inhibition, one must preexpose one group of subjects to the stimulus of interest, while not giving such stimulus preexposure to a control group. In the test phase, both groups must learn to form an association between that stimulus and a new event. When the stimulus-preexposed group learns the new association to that stimulus more poorly than does the control group, we say that latent inhibition has been demonstrated.

In general, the effects of organismic variables have not been subjects of investigation in regard to latent inhibition. The few exceptions include the variables of age, sex, and handling, with the former receiving the greatest attention by far. These areas are represented by a small number of experimental investigations, and some conclusions may be reached from the controlled comparisons. In addition, however, there is a body of information that is derivable from the fact that experiments on latent inhibition have been performed in a variety of species. It is therefore possible to examine, between experiments, effects that, with considerable caution, may be attributable to species differences. Prudence is dictated by the fact that across studies there is almost complete confounding of species and testing procedures.

Age

Only two studies have looked at the effect of age on latent inhibition within the conditioned suppression paradigm (Cone, 1974; Wilson & Riccio, 1973). This, of course, is not surprising, because the basic technique, whether tube licking or bar pressing, requires a relatively mature motor system. Other procedures, such as odor preference, are more amenable to dealing with the inherent problems of testing immature organisms.

In a sketchy report that is difficult to evaluate, Cone (1974) reported an age-related latent inhibition effect in a one-trial conditioned suppression test. With rats that ranged in age from 30 to 365 days, all groups showed latent inhibition to a light CS, except the 90–120 day-old group.

In general, conditioned attention theory, CAT (Lubow, Schnur, & Rifkin, 1976; Lubow, Weiner, & Schnur, 1981), states that nonreinforced preexposure to a stimulus retards subsequent conditioning to that stimulus because during such preexposure the animal learns not to attend to it. The theory is based on the use of attention as a hypothetical construct, with the characteristics of a Pavlovian response, and on the specification of reinforcement conditions that modify attention.

The assumption that changes in attention to stimuli are a function of reinforcement conditions may be traced to Lashley (1929) and Krechevsky (1932). Likewise, Lawrence (1949) suggested that the “acquired distinctiveness of cues” might be a gradual learning process subject to traditional analysis. In more recent theorizing, changes in attention as a function of reinforcement conditions have been emphasized by Mackintosh (1975), Frey and Sears (1978), and Pearce and Hall (1980), as well as in the “selective attention” theories of Lovejoy (1968), Sutherland and Mackintosh (1971), Trabasso and Bower (1968), and Zeaman and House (1963).

In similarity to selective attention theories, CAT treats attention as a response, occurring on stimulus presentation, the probability of which is increased when it is followed by reinforcement and decreased when it is not reinforced. However, CAT differs from those theories in a number of important respects: the conditions specified for the changes in the attentional response; the mechanism postulated to govern such changes; and the course of these changes with repeated stimulus presentation.

Originally, this book was to have been organized such that a separate chapter would be devoted to each of the major paradigms within which latent inhibition had been demonstrated. Although such a categorization proved to be convenient for the initial organizing of the hundreds of published reports, it soon became clear that there was an absence of major differences in the latent inhibition phenomena that could be attributed to the type of testing procedure employed. Therefore, it was decided to present the empirical evidence for latent inhibition on the basis of manipulated variables, organismic variables, and variations across experiments. It is, of course, in this latter category that testing procedures are to be found. Procedures are placed at the head of the list, for two reasons: first, to present evidence that latent inhibition is indeed a ubiquitous phenomenon, found across a very wide range of testing procedures; second, by accomplishing the preceding, to avoid, in subsequent sections, the necessity of repetitiously labeling the experimental procedures for each point that is made.

Before identifying those paradigms in which latent inhibition has been investigated, two general comments are in order, one concerning the number of stages in the procedure, and the second concerning the problem of differentiating the conditioned response (CR) to the target stimulus in the test from the unconditioned response (UR) to that stimulus.

Although conditioned attention theory was developed specifically to account for latent inhibition effects, it is also relevant to other phenomena in the area of learning, particularly to those situations where the target stimulus is presented under conditions in which attention is diverted from it by competing stimuli, such as in blocking and overshadowing, or in which stimuli are presented repeatedly, as S1 in the preexposure phase of the learned helplessness paradigm, or S1–S2 in the first phase of sensory preconditioning. In regard to the first of these areas, Lubow, Weiner, and Schnur (1981) have already commented on the implications of conditioned attention theory for understanding blocking and overshadowing, as well as the feature positive effect. Similarly, in the second area, the relationships among CAT, learned helplessness, habituation, and sensory preconditioning have been explored. With the exception of learned helplessness, there is little to be added to the already published analysis, and the interested reader is referred to the original source (Lubow, Weiner, & Schnur, 1981). However, in regard to learned helplessness, new and important materials are available, enough to warrant a reconsideration of the relationship between latent inhibition and learned helplessness. In addition, a literature has recently developed on the relationship between latent inhibition, as it reflects normal attentional processes, and schizophrenia, often regarded as characterized by attentional dysfunction. These two topics, learned helplessness and schizophrenia, will be explored in the following two sections.

In the past, the extensive efforts to understand latent inhibition were directed at behavioral analyses of the phenomenon, the data from which we have discussed at some length in the previous chapters. More recently, however, a keen interest has developed in the neural substrates of latent inhibition. Although this certainly reflects a general trend in the experimental psychology of learning, the additional focus also can be partially explained by the relatively new interest in the area of attention among psychologists, as well as by the consensual opinion that latent inhibition reflects some aspect of attention (e.g., Lubow, Weiner, & Feldon, 1982). More specific manifestations of this direction can be seen in various attempts to evaluate the attentional deficits of schizophrenia by assessing impairments of latent inhibition (Baruch et al., 1988a,b; Lubow et al., 1987) and, at the same time, to explore the animal amphetamines model of schizophrenia by examining the effects of amphetamines as well as neuroleptics on latent inhibition in animals (e.g., Solomon et al., 1981; Solomon & Staton, 1982; Weiner et al., 1984; Weiner & Feldon, 1987; Weiner, Feldon, & Katz, 1987; Weiner, Lubow, & Feldon, 1981). In addition to these efforts, which will be described in the section on the effects of dopaminergic manipulations on latent inhibition, other brain systems have also been studied in this regard. These will be reviewed and discussed separately in sections on noradrenergic, serotonergic, cholinergic, septo-hippocampal, and opiate manipulations.

In the preceding chapter we developed and applied CAT to the data from latent inhibition studies with animals, but we ignored the data from experiments with human subjects. It will be recalled that these human studies present a pattern of results not entirely consistent with that obtained from lower organisms. To review, the major findings, presented in chapter 5, are as follows:

1. There are standard stimulus preexposure procedures that will produce latent inhibition in young children, but not in older children or adults.

2. If these same procedures are coupled with a masking task, then latent inhibition can be produced in older children and adults.

3. Nevertheless, there are some procedures (i.e., electrodermal conditioning, conditioned tasted aversion, and Ivanov-Smolensky conditioning) that may not require masking for the production of latent inhibition.

The last point will be dismissed, perhaps somewhat cavalierly, because of the difficulty in separating the unconditioned response to the CS (orienting or otherwise) from the conditioned response to the CS in the test phase. As a result, one cannot determine whether stimulus preexposure produces an artifactual reduction in the conditioned response or a reduction in the associability of the CS. This, it will be recalled, is similar to the problem encountered in a number of animal conditioning paradigms. In addition, these studies with humans may have included unplanned masking procedures – either by the very nature of the procedure or by instruction.

Let us now consider the effects of age and masking on latent inhibition in humans.