Artificial grammar learning meets formal language theory: an overview

Formal language theory (FLT), part of the broader mathematical theory of computation, provides a systematic terminology and set of conventions for describing rules and the structures they generate, along with a rich body of discoveries and theorems concerning generative rule systems. Despite its name, FLT is not limited to human language, but is equally applicable to computer programs, music, visual patterns, animal vocalizations, RNA structure and even dance. In the last decade, this theory has been profitably used to frame hypotheses and to design brain imaging and animal-learning experiments, mostly using the 'artificial grammar-learning' paradigm. We offer a brief, non-technical introduction to FLT and then a more detailed analysis of empirical research based on this theory. We suggest that progress has been hampered by a pervasive conflation of distinct issues, including hierarchy, dependency, complexity and recursion. We offer clarifications of several relevant hypotheses and the experimental designs necessary to test them. We finally review the recent brain imaging literature, using formal languages, identifying areas of convergence and outstanding debates. We conclude that FLT has much to offer scientists who are interested in rigorous empirical investigations of human cognition from a neuroscientific and comparative perspective.     

Risk, unexpected uncertainty, and estimation uncertainty: Bayesian learning in unstable settings

Recently, evidence has emerged that humans approach learning using Bayesian updating rather than (model-free) reinforcement algorithms in a six-arm restless bandit problem. Here, we investigate what this implies for human appreciation of uncertainty. In our task, a Bayesian learner distinguishes three equally salient levels of uncertainty. First, the Bayesian perceives irreducible uncertainty or risk: even knowing the payoff probabilities of a given arm, the outcome remains uncertain. Second, there is (parameter) estimation uncertainty or ambiguity: payoff probabilities are unknown and need to be estimated. Third, the outcome probabilities of the arms change: the sudden jumps are referred to as unexpected uncertainty. We document how the three levels of uncertainty evolved during the course of our experiment and how it affected the learning rate. We then zoom in on estimation uncertainty, which has been suggested to be a driving force in exploration, in spite of evidence of widespread aversion to ambiguity. Our data corroborate the latter. We discuss neural evidence that foreshadowed the ability of humans to distinguish between the three levels of uncertainty. Finally, we investigate the boundaries of human capacity to implement Bayesian learning. We repeat the experiment with different instructions, reflecting varying levels of structural uncertainty. Under this fourth notion of uncertainty, choices were no better explained by Bayesian updating than by (model-free) reinforcement learning. Exit questionnaires revealed that participants remained unaware of the presence of unexpected uncertainty and failed to acquire the right model with which to implement Bayesian updating.     

Mechanisms of stimulation of olfactory neurons: an essay

It is widely believed that the mechanism by which olfactoryneurons are stimulated by odorants is via specific receptorproteins located on or in the apical cell membranes of the neurons.In this review alternative mechanisms are explored and a modelis presented in which sensitivity and specificity can be accountedfor by the general irritability of cells to chemicals. The premiseson which the model is based are: (i) Groups of olfactory nerves fire action potentials whichare summated at the first synapse, in the olfactory bulb. Theseneurons fire spontaneously because their resting potentialsare unstable. (ii) What we call olfactory quality is encoded in patterns offiring of different olfactory neurons. (iii) Olfactory neurons are not identical. The sources of theirindividuality include differences in their age, spatial locationin the plane of the tissue, depth within the tissue and distancefrom the nearest capillary. The heterogeneity will be reflectedas differences in their metabolism and chemical composition. (iv) Most of the surface area of olfactory neurons is bathedin mucus that is secreted by Bowman's glands and sustentacularcells. Therefore, the resting potential and firing frequencyof each will depend upon the composition of the mucus. It follows from these premises that alterations in the mucuscomposition, extracellular fluid composition, metabolic stateor properties of neuronal membranes will alter the resting potentialand, therefore, the firing frequency of each neuron. The effectswill vary among neurons because of their individuality evenif the stimulus is distributed uniformly across the olfactorymucosa, different regions will be exposed to different patternsof stimulus concentration because odorants are differentiallyadsorbed and metabolized. According to this scheme qualitativelydifferent patterns of neural activity will result from differentodorants, the firing of some neurons being inhibited and thatof others being stimulated in each case. This will be true independentlyof the existence of specific receptor proteins, the effectsof which must be superimposed upon the general effects of nonspecificirritability.     

Future oriented conservation: knowledge governance,uncertainty and learning

Despite significant progress in understanding climate risks, adaptation efforts in biodiversity conservation remain limited. Adaptation requires addressing immediate conservation threats while also attending to long term, highly uncertain and potentially transformative future changes. To date, conservation research has focused more on projecting climate impacts and identifying possible strategies, rather than understanding how governance enables or constrains adaptation actions. We outline an approach to future-oriented conservation that combines the capacities to anticipate future ecological change; to understand the implications of that change for social, political and ecological values; and the ability to engage with the governance (and politics) of adaptation. Our approach builds on the adaptive management and governance literature, however we explicitly address the (often contested) rules, knowledge and values that enable or constrain adaptation. We call for a broader focus that extends beyond technical approaches to acknowledge the socio-political challenges inherent to adaptation. More importantly, we suggest that conservation policy makers and practitioners can use this approach to facilitate learning and adaptation in the context of complexity, transformational change and uncertainty.     

Unfinished business: an essay on finally leaving the bench

Spontaneous and induced chromosome aberrations have been studied over more than a century. The resolution of detection of aberrations has depended on the improvement of available techniques. An overview on the major high lights in this area of research, from the time of solid staining to fluorescence in situ hybridization technique is presented in this review.     

Knowing human moral knowledge to be true: an essay on intellectual conviction

The question addressed in this article is how people come to know the foundational axioms of their moral systems as true and correct. Drawing on my fieldwork among the Himba of northwestern Namibia, I argue that the most potent form of intellectual conviction is not generated through the external manipulations of ritual, but through a deeply internal experience in which moral knowledge coalesces with a subjectively perceived experience of timeless universality.     

Transductions to generate plant form and pattern: an essay on cause and effect

Many complex processes can be broken into transduction steps where one state is converted to another by a well defined activity. One difficulty for analysis is that transductions occur in chains or networks. Another, of primary concern here, is that a single transduction can be complex. Some such transductions can efficiently explain phenomena often thought to be summations or orchestrations of many simple transductions. Pattern formation is in this category. For a wide range of transductions one can define cause and effect in a differential equation. In its integral one can define the before and after states. The main experimental tactic to characterize unknown transductions is co-variation. The before state (input) is altered, change in the after state (output) is assayed. Thus an unknown transduction, with cause and effect embodied in the differential, is investigated through long-term changes in its integral. This is fully practical when all of the integral is known or readily surmised, as in simple discrete biochemical transductions. As causal differential expressions become complex, their integrals become more versatile in generating output because this changes not only with variation in the expression itself but also with boundary conditions and limits. These very features, however, make such a function increasingly intractable to discovery by co-variation. Only a small part of the integral is embodied in the before and after states; the remainder is not readily surmised. Accordingly, in contrast to reliance on the role of controls to deduce unknown simple transductions, the complex ones are generally established through formalization of the differential nature of the process itself.