Neuro-psychoanalysis: a new revolution?

In this brief article we will explain the current connections between psychoanalytic concepts, like Ego or Id, and Neurology and the computational study of the episodic memory for artificial agents. We will refer to the book by Solms and Turnbull, "The Brain and the Inner World" and to the recent work by Andreas Gruber at the Vienna University.

In 1923, Freud recognized that the rational part of the mind is not necessarily conscious: consciousness was not a fundamental organizing principle of the architecture of the mind. Freud attributed the functional properties previously assigned to the system Conscious-Preconscious to the "Ego", being conscious only a small portion of the Ego´s activities. Its main property was the capacity for inhibition. Freud thought of the capacity to inhibit drive energies, the basis of all the Ego´s rational. Consciousness over time allows the development of what Damasio (1999) calls an "autobiographical self". In psychoanalytic terms, the core "self" might be described as a perception of the current state of the "Id", whereas the autobiographical self is synonymous with the "Ego". Episodic memory is the fundamental element of the "autobiographical self". Following to Schacter (1996), the episodic memory system allows explicitly to recall the personal incidents that define the lives of human beings. Very recently, Andreas Gruber (2007) at the Vienna University, has built a computational model of episodic memory that is based on Freud´s Ego-Superego-Id personality model. A pre-decision module, representing the Freudian "Id", consists of the drives and the basic emotions modules. In this model, low-level decision making is done. If a drive is very high, the system will try to bring the drive back into its balanced range. By the other side, perceived situations are handed over to the the high-level decision making module: it represents the "Ego". Decision module interacts with the episodic memory by searching for similar situations to the current one including their emotional rating. A behavior may be triggered directly by a complex emotion while reactive responses arise from the pre-decision module, whereas routines are longer sequences of actions stored in the procedural memory.

It is very curious to verify that the Freudian paradigm, always impugned by the absence of scientific methodological rigour, currently is guiding the search in Neurology and Artificial Intelligence. Are we witnessing a new and revolutionary synthesis?

Dunlosky & Metcalfe: "Metacognition"

Two prestigious authors, John Dunlosky and Janet Metcalfe, have published on September the first introduction to Metacognition for undergraduates. Dunlosky is a Professor of Psychology at Kent State University and Metcalfe is a Professor of Psychology and of Neurobiology and Behavior at Columbia University. This book is an excellent tool for understanding in a concise way the state of the art of the emergent field of the Metacognition. From a didactic point of view, we remark the presence in each chapter of boxes presenting specific subjects and closing the chapters, questions discussion and review concepts which allow to consolidate the learning.

An introduction starts presenting the metacognitive model of Nelson and Narens (1990), the most followed model in this area. In this model, the interplay between the meta-levels and the object-level defines the two process-based activities of metacognition: monitoring and control: control is exerted whenever the meta-level modifies the object-level; controlling the object-level provides no information about the states of the object-level. For it, you must monitor the object-level activities so that you can update the model of them.

Chapter 2 presents a very interesting History about the theoretical development of the concept. Perhaps it´s the first historical compendium published. Since the modern pioneer John Flavell to Asher Koriat many names and photos of researchers are included.

Section 1 analyzes the main empirical methods for investigating the Metacognition and chapter 4 speaks about feeling-of-Knowing (FOK) judgments and Tip-of-the-Tongue states. In the first case, Dunlosky and Metcalfe introduce theories like "Target Stregth" (Hart, 1967), "Cue Familiarity" (Reder, 1987) or "Target Accessibility" (Koriat, 1993). An interesting section about the contributions by Lynne Reder and colleagues on strategy selection by means of (FOK) judgments, finishes the chapter.

Judgments of learning (JOLs) are decisive for Education. In a seminal work, Arbuckle and Cuddy (1969) argued that if paired associates "differed in associative strengths immediately following presentation, subjects should be able to detect these differences just as they can detect differences in strength of any other input signal". Their students studied short lists of paired associates, and inmediately after studying each one, they judged whether or not they would recall it. Following Dunlosky and Metcalfe (p. 94), research on JOLs has focused on contexts in which extra study typically does improve memory. Besides self-feedback from intervening test trials does benefit the relative accuracy of JOLs. There are several hypothesis explaning the effects of JOLs. Begg et al. (1989) proposed that JOLs are based on the ease of processing an item inmediately prior to making the judgment. Bjork and Schwartz (1998) have demonstrated the effect of retrieval fluency on JOLs. Koriat (1997) has used the cue-utilization approach. Acording to this perspective, people use a variety of cues-item relatedness, number of study trials...-to infer whether an item will be remembered.

Confidence judgments or retrospective confidence (RC) judgments are analysed in chapter six. What causes the overconfidence effect? Perhaps cognitive biases distort people´s judgments but the overconfidence effect is probably an artifact of experimental methods.

Source judgments (chapter seven) involve remembering the source of a memory or the context in which a memory originally occurred. Between the factors that influence source-monitoring accuracy are the similarity of the sources and emotion and imagery.

Chapter eight is a very interesting one. The authors review some findings on people´s assessment of the truth of witnesses´memories based on their expressed confidence, and on people´s abilities to detect lies. Many cases are presented specially about hindsight bias: this effect has important implications for the criminal justice system, for isues as whether jurors are able to disregard testimony that has been ruled inadmisible.

In chapter nine, are considered in detail the wide connections of Metacognition to Education and this textbook finishes including a section about life-span development. We remark in chapter ten the interesting references on development of theory of mind and metacognition. In the other side (chapter eleven), it´s reviewed the problem of aging and memory monitoring.

This book is not only the first textbook to focus on Metacognition but perhaps the first book in covering metacognitive research in an unified framework. It´s also an excellent handbook for more advanced students and has been writen by two authentic champions of the subject.