I Introduction
The prefrontal cortex is the cortex of the anterior pole of the mammalian brain. In characterizing the anterior part of the frontal lobe with the adjective prefrontal, we make loose, if not improper, use of the prefix pre (literally we place that cortex in midair!). Nevertheless, that designation has been condoned by so much usage that it seems unwarranted to discard it for semantic reasons.
Here, the prefrontal cortex is defined as the part of the cerebral cortex that receives projections from the mediodorsal nucleus of the thalamus. This anatomical definition is applicable to all mammalian brains. It takes into consideration the possibility that the relationship with a well-defined thalamic nucleus reflects an identifiable function or group of functions. Of course, such reasoning is based on analogy with specific thalamic nuclei and their cortical projection areas, an analogy that may not be entirely appropriate. Furthermore, the functions of the mediodorsal nucleus are not well known, and the prefrontal cortex is also connected to many other cerebral structures.
On the other hand, the definition by relationship has here the merit of obeying the reasonable principle that the physiology of a cortical region can be meaningfully studied and understood only in the context of its anatomical connections with other structures (Creutzfeldt, 1977). In this respect, however, the connectivity of the prefrontal cortex with other parts of the cortex may be more important than its thalamic connectivity.
The basic function of the prefrontal cortex is the representation and execution of new forms of organized goal-directed action. All the so-called executive functions of the prefrontal cortex serve that superordinate function in one way or another.
The goals of an organism, especially the human organism, can vary immensely, and so do the timescale and means to achieve them. Also variable are the motives for action and the emotions that accompany it, as well as their influence at any step in the pursuit of a goal. Depending on these factors, each of the executive functions of the prefrontal cortex may be brought into play at one time or another. In the human and non-human primate, as in other animals, each one of these functions has a dominant if not specific regional location in the prefrontal cortex. In any event, the diversity of regional commitments, as well as of the connectivity of different prefrontal areas, has to be analyzed in depth, because therein lies the key to their function. We shall never understand fully the functions of the prefrontal cortex if we neglect the operations of its components.
At the same time, we must keep in mind the wider structural and dynamic context in which those operations take place. This context is defined by two basic biological principles that set the background for any discussion of this cortex. One is the evolutionary hierarchy of cortical and subcortical structures dedicated to the organization of goal-directed actions. The other is the perceptionâaction cycle; that is, the cybernetic circulation of information processing that governs the interactions of the organism with its environment. Both principles are structurally and dynamically intertwined.
The first, the hierarchical vision of the neuroscience of action, has its origin in the writings of John Hughlings Jackson (1958), a scholarly physician who practiced neurology in Londonâs Queen Square Hospital at the end of the nineteenth and beginning of the twentieth century. Based on his studies of motor functions and their disorders, Jackson advanced the idea that the structures of the central nervous system, its motor structures in particular, were hierarchically organized in the order determined by evolution: structures representing and coordinating simple movements at the bottom (basal ganglia, pyramidal system, cerebellum), and those representing and coordinating new complex behavior at the top (prefrontal and premotor cortices). As clinical observations clearly demonstrate, lesions at a particular level of the motor hierarchy lead to paralyses of movements organized at that level and, at the same time, to the release of simpler, automatic movements from lower levels of the hierarchy. Jackson characterized such a pathological disorder as âdissolution,â a term he coined as an opposite to evolution, since the disorder indicated that upon failure of higher levels of the hierarchy, the nervous system regressed to the performance of movements that are more primitive from an evolutionary point of view.
An evident corollary of Jacksonâs theory, which he drew quite early (Jackson, 1882), is that the evolutionarily lower structures and their functions are nested under the higher structures and functions, which they normally serve. When the higher ones fail, the lower are released from their control. This is true in all cortical hierarchies of action, but is most obvious in the hierarchy of areas dedicated to the phylogenetically most advanced cognitive activity, the spoken language. At the lowest cortical level of the speech hierarchy is the sensorimotor cortex, which controls the representation and articulation of simple speech utterances. By high-resolution methods in humans, precise temporal patterns can be recorded in that cortex that correspond to the successive activations of oral and laryngeal muscles during the articulation of vowels and consonants (Bouchard et al., 2013). Words are made of phonemes and morphemes, which are organized into sentences in hierarchically higher cortical regions, such as the premotor cortex and Brocaâs area.
As a result of his clinical research, Jackson reserved for the prefrontal cortex the representation and organization of what he called âpropositionalâ language. It should not escape us that, in logic and linguistics, the term propositional implies novelty, complexity, and even a future dimension â which all proposals have. At the same time, these are the characteristics that make language a uniquely human activity (Berwick et al., 2013) and place its most novel and complex aspects in the prefrontal cortex, at the summit of the evolutionary hierarchy of neural structures for action.
The concept of the perceptionâaction cycle also has a deep root in biology. Prefrontal areas, networks, and functions are not simply interdependent; they are cooperative. The temporal organization of complex and novel actions toward their goal is the product of the dynamics of the perceptionâaction cycle, which consists of the coordinated participation of neural structures in the successive interactions of the organism with its environment in the pursuit of a goal. Thus, the perceptionâaction cycle is the cortical substrate for the processing of information between the organism and its environment; the prefrontal cortex constitutes the highest stage of neural integration in that cycle. In the course of a goal-directed sequence of actions, signals from the internal milieu and the external environment are processed through hierarchically organized neural channels and lead into the prefrontal cortex (internal signals into orbitomedial, external signals into lateral prefrontal cortex). There, the signals generate or modulate further action, which in turn causes changes in the internal and external environments, changes which enter the processing cycle toward further action, and so on until the goal is reached. At each hierarchical level of the cycle, there is feedback to prior levels. At the highest level, there is re-entrant feedback from the prefrontal cortex to the posterior association cortex, which plays a critical role in working memory, set, and monitoring.
Those two general concepts, hierarchy and the perceptionâaction cycle, mark the theoretical backdrop in the cerebral cortex at large against which the functions of the prefrontal cortex must be viewed. Both cortices, posterior and frontal, are hierarchically organized. Whereas the posterior cortex is devoted to perceptual and mnemonic functions, the entirety of the frontal cortex, including its prefrontal region, is devoted to action of one kind or another, whether it is skeletal movement, ocular movement, the expression of emotion, speech, or visceral control. The action can even be mental and internal, such as reasoning. The frontal cortex is therefore âdoerâ cortex, much as the posterior cortex is âsensorâ cortex (both reflecting up in the cortex the polarity of functions existing in the anterior and posterior horns of the spinal cord). In sum, the posterior cortex and the frontal cortex constitute the cortical infrastructure for the perceptionâaction cycle.
The frontal cortex does nothing by itself. It works in the perceptionâaction cycle with other cortices, with subcortical structures, and with certain sectors of the sensory and motor apparatus and of the autonomic system. There is, however, considerable specialization of action within it. Accordingly, there are frontal areas for eye movement, for skeletal movement of various body parts, for speech, for emotional expression, and so on. More importantly in what concerns us here, the specialized areas within the prefrontal cortex, whatever the action domain they represent, contribute their share to the common cognitive and emotional functions that drive the neocortex as a whole. Those functions are essentially integrative and goal directed. They are also, as we will see, new for the organism; they are new as that organism has to meet new circumstances, now or in the future, and has to adapt to them. In that sense, the prefrontal cortex is not only adaptive, but also preadaptive.
As organisms evolve, their actions become more complex and idiosyncratic, their goals more remote in space and time, and their reasons or motives for attaining them less transparent, more based on probability and prior experience than on peremptory instinctual need. Furthermore, action in general becomes more deliberate and voluntary. With this evolution of biological action, and presumably because of it, the most anterior sector of the frontal cortex, which we call the prefrontal cortex, grows substantially â in relative size â as evolution progresses, and so does its functional role. Its growth reaches a maximum in the human primate. The prefrontal cortex of the lateral or outer frontal convexity, which is essential for cognitive functions and intelligent behavior, undergoes greater development than that of the medial and inferior (orbital) surfaces, which are critically involved in emotional behavior. Although their functions are interdependent and integrated in the behavior of the organism, lateral and orbitomedial cortices require somewhat different methodologies for their study.
In this book, we shall examine the prefrontal cortex by systematically reviewing data from each of the contributing methodologies. As we proceed from the basic facts of anatomy to neuropsychology, to neurophysiology, and to neuroimaging, my own conceptual point of view will become progressively more explicit. This introduction outlines it in broad strokes and the last chapter describes it in detail.
In the 33 years since the first edition of this book, my theoretical position on the prefrontal cortex has changed considerably as new facts have demanded it, but some of the basic elements of my initial view are still valid. To begin with, there is now wide agreement with the concept I held then, that the lateral prefrontal cortex is critical for the cognitive functions that mediate the temporal organization of actions. These functions include planning, decision-making, and topâdown attention, the latter with its three subcomponents of working memory, set, and inhibitory control. Another surviving view is that the limbic, âdysgranular,â cingular, medial, and orbital areas of the prefrontal cortex, while also involved in those functions, modulate their emotional and affective components. Those areas can even initiate goal-directed actions in the emotional perceptionâaction cycle, which runs parallel to the cognitive one and interacts with it â in orbital prefrontal cortex. We also know, as we knew then, that two specialized regions at the transition between prefrontal and premotor cortex serve the coordination of eye movements (area 8) and speech (areas 44â45, Brocaâs area).
With continuing research, other views on the prefrontal cortex have emerged in later years, which, without substantially modifying the previous ones, add to them two essential accents: one on novelty and the other on the future. Any series of purposive actions that is new and thus deviates from rehearsed or automatic routine or instinctual order necessitates the lateral prefrontal cortex. The longer the series, and therefore the further it extends into the future, the greater the need for that cortex. Time is only one factor, however, among those determining that need; other factors include the complexity of the actions and of the information on which they are based, and still another the uncertainties or ambiguities in that information. There is considerable trade-off between those factors. For example, a monkey with a prefrontal deficit may fail at a simple and thoroughly rehearsed task, such as delayed response, not only because of the interval of time between cue and response, but also because of the competitive interference â a source of uncertainty and ambiguity â between two alternative cues that succeed each other at random from one trial to the next. Still, the question may be asked, what is new for the prefrontal cortex in a task as stereotypical as that one? The answer to that question is that the cue for every trial, although part of an old repertoire, is unpredictable and new for that trial. This critical element of built-in novelty is ignored by many studies using delay tasks.
Yet time is probably the single most important attribute placing a complex and novel sequence of behavior under the physiological purview of the lateral prefrontal cortex (Fuster, 2001). Only this part of the cerebral cortex can provide that âtemporal gestaltâ with the coherence and coordination of actions that are essential for the organism to reach its new goal. Both coherence and coordination derive from the capacity of the prefrontal cortex to organize new goal-directed actions in the time domain, which in my view is the most general and characteristic of all prefrontal functions in the primate. The importance of this temporal-organizing function in mammalian behavior cannot be overstated. Without it, there is no execution of novel, elaborate behavior, no speech fluency, no higher reasoning, and no creative activity with more than a minimal temporal dimension; only temporal concreteness is left, the here and now.
For it to function properly, however, the overarching prefrontal function of organizing new goal-directed actions in time necessitates two essential elements: (1) a cortex-wide infrastructure for the representation of knowledge and memory in the form of distributed cognitive networks, which I have called cognits (Fuster, 2009); and (2) a number of executive integrative functions that will manipulate those cognits in temporal integration. There are essentially five of these functions: planning, decision-making, working memory, preparatory set, and inhibitory control. Note that they all have a future perspective and that they interact with one another in the organization of goal-directed actions.
Indeed, all the cognitive functions of the cortex take place on a neural substrate of neural representation. That substrate is made of a vast neuronal network that is the repository of permanent, though modifiable, long-term memory, and knowledge (knowledge is semantic memory). That substrate extends to the hippocampus, which is a portion of ancient cortex that is essential for the consolidation of all explicit/declarative memory and knowledge. Individual memories or cognits are subcomponents, also network-like,...