Prefrontal function and cognitive control: from action to language

A deep hierarchy of predictions enables online meaning extraction in a computational model of human speech comprehension

Understanding speech requires mapping fleeting and often ambiguous soundwaves to meaning. While humans are known to exploit their capacity to contextualize to facilitate this process, how internal knowledge is deployed online remains an open question. Here, we present a model that extracts multiple levels of information from continuous speech online. The model applies linguistic and nonlinguistic knowledge to speech processing, by periodically generating top-down predictions and incorporating bottom-up incoming evidence in a nested temporal hierarchy. We show that a nonlinguistic context level provides semantic predictions informed by sensory inputs, which are crucial for disambiguating among multiple meanings of the same word. The explicit knowledge hierarchy of the model enables a more holistic account of the neurophysiological responses to speech compared to using lexical predictions generated by a neural network language model (GPT-2). We also show that hierarchical predictions reduce peripheral processing via minimizing uncertainty and prediction error. With this proof-of-concept model, we demonstrate that the deployment of hierarchical predictions is a possible strategy for the brain to dynamically utilize structured knowledge and make sense of the speech input.

Conscious cognitive effort in cognitive control

Cognitive effort is thought to be familiar in everyday life, ubiquitous across multiple variations of task and circumstance, and integral to cost/benefit computations that are themselves central to the proper functioning of cognitive control. In particular, cognitive effort is thought to be closely related to the assessment of cognitive control's costs. I argue here that the construct of cognitive effort, as it is deployed in cognitive psychology and neuroscience, is problematically unclear. The result is that talk of cognitive effort may paper over significant disagreement regarding the nature of cognitive effort, and its key functions for cognitive control. I highlight key points of disagreement, and several open questions regarding what causes cognitive effort, what cognitive effort represents, cognitive effort's relationship to action, and cognitive effort's relationship to consciousness. I also suggest that pluralism about cognitive effort-that cognitive effort may manifest as a range of intentional or non-intentional actions the function of which is to promote greater success at paradigmatic cognitive control tasks-may be a fruitful and irenic way to conceive of cognitive effort. Finally, I suggest that recent trends in work on cognitive control suggests that we might fruitfully conceive of cognitive effort as one key node in a complex network of mental value, and that studying this complex network may illuminate the nature of cognitive control, and the role of consciousness in cognitive control's proper functioning. This article is categorized under: Philosophy > Consciousness Philosophy > Psychological Capacities Neuroscience > Cognition.

Curiosity as a metacognitive feeling

Curious information-seeking is known to be a key driver for learning, but characterizing this important psychological phenomenon remains a challenge. In this article, we argue that solving this challenge requires qualifying the relationships between metacognition and curiosity. The idea that curiosity is a metacognitive competence has been resisted: researchers have assumed both that young children and non-human animals can be genuinely curious, and that metacognition requires conceptual and culturally situated resources that are unavailable to young children and non-human animals. Here, we argue that this resistance is unwarranted given accumulating evidence that metacognition can be deployed procedurally, and we defend the view that curiosity is a metacognitive feeling. Our metacognitive view singles out two monitoring steps as a triggering condition for curiosity: evaluating one's own informational needs, and predicting the likelihood that explorations of the proximate environment afford significant information gains. We review empirical evidence and computational models of curiosity, and show that they fit well with this metacognitive account, while on the contrary, they remain difficult to explain by a competing account according to which curiosity is a basic attitude of questioning. Finally, we propose a new way to construe the relationships between curiosity and the human-specific communicative practice of questioning, discuss the issue of how children may learn to express their curiosity through interactions with others, and conclude by briefly exploring the implications of our proposal for educational practices.

Lateral Prefrontal Cortex Is a Hub for Music Production from Structural Rules to Movements

Complex sequential behaviors, such as speaking or playing music, entail flexible rule-based chaining of single acts. However, it remains unclear how the brain translates abstract structural rules into movements. We combined music production with multimodal neuroimaging to dissociate high-level structural and low-level motor planning. Pianists played novel musical chord sequences on a muted MR-compatible piano by imitating a model hand on screen. Chord sequences were manipulated in terms of musical harmony and context length to assess structural planning, and in terms of fingers used for playing to assess motor planning. A model of probabilistic sequence processing confirmed temporally extended dependencies between chords, as opposed to local dependencies between movements. Violations of structural plans activated the left inferior frontal and middle temporal gyrus, and the fractional anisotropy of the ventral pathway connecting these two regions positively predicted behavioral measures of structural planning. A bilateral frontoparietal network was instead activated by violations of motor plans. Both structural and motor networks converged in lateral prefrontal cortex, with anterior regions contributing to musical structure building, and posterior areas to movement planning. These results establish a promising approach to study sequence production at different levels of action representation.

Hierarchical control as a shared neurocognitive mechanism for language and music

Although comparative research has made substantial progress in clarifying the relationship between language and music as neurocognitive systems from both a theoretical and empirical perspective, there is still no consensus about which mechanisms, if any, are shared and how they bring about different neurocognitive systems. In this paper, we tackle these two questions by focusing on hierarchical control as a neurocognitive mechanism underlying syntax in language and music. We put forward the Coordinated Hierarchical Control (CHC) hypothesis: linguistic and musical syntax rely on hierarchical control, but engage this shared mechanism differently depending on the current control demand. While linguistic syntax preferably engages the abstract rule-based control circuit, musical syntax rather employs the coordination of the abstract rule-based and the more concrete motor-based control circuits. We provide evidence for our hypothesis by reviewing neuroimaging as well as neuropsychological studies on linguistic and musical syntax. The CHC hypothesis makes a set of novel testable predictions to guide future work on the relationship between language and music.

The Role of the Dorsolateral Prefrontal Cortex for Speech and Language Processing

This review article summarizes various functions of the dorsolateral prefrontal cortex (DLPFC) that are related to language processing. To this end, its connectivity with the left-dominant perisylvian language network was considered, as well as its interaction with other functional networks that, directly or indirectly, contribute to language processing. Language-related functions of the DLPFC comprise various aspects of pragmatic processing such as discourse management, integration of prosody, interpretation of nonliteral meanings, inference making, ambiguity resolution, and error repair. Neurophysiologically, the DLPFC seems to be a key region for implementing functional connectivity between the language network and other functional networks, including cortico-cortical as well as subcortical circuits. Considering clinical aspects, damage to the DLPFC causes psychiatric communication deficits rather than typical aphasic language syndromes. Although the number of well-controlled studies on DLPFC language functions is still limited, the DLPFC might be an important target region for the treatment of pragmatic language disorders.

Brain-inspired model for early vocal learning and correspondence matching using free-energy optimization

We propose a developmental model inspired by the cortico-basal system (CX-BG) for vocal learning in babies and for solving the correspondence mismatch problem they face when they hear unfamiliar voices, with different tones and pitches. This model is based on the neural architecture INFERNO standing for Iterative Free-Energy Optimization of Recurrent Neural Networks. Free-energy minimization is used for rapidly exploring, selecting and learning the optimal choices of actions to perform (eg sound production) in order to reproduce and control as accurately as possible the spike trains representing desired perceptions (eg sound categories). We detail in this paper the CX-BG system responsible for linking causally the sound and motor primitives at the order of a few milliseconds. Two experiments performed with a small and a large audio database show the capabilities of exploration, generalization and robustness to noise of our neural architecture in retrieving audio primitives during vocal learning and during acoustic matching with unheared voices (different genders and tones).

How Abstract (Non-embodied) Linguistic Representations Augment Cognitive Control

Recent scholarship emphasizes the scaffolding role of language for cognition. Language, it is claimed, is a cognition-enhancing niche (Clark, 2006), a programming tool for cognition (Lupyan and Bergen, 2016), even neuroenhancement (Dove, 2019) and augments cognitive functions such as memory, categorization, cognitive control, and meta-cognitive abilities ("thinking about thinking"). Yet, the notion that language enhances or augments cognition, and in particular, cognitive control does not easily fit in with embodied approaches to language processing, or so we will argue. Accounts aiming to explain how language enhances various cognitive functions often employ a notion of abstract representation. Yet, embodied approaches to language processing have it that language processing crucially, according to some accounts even exclusively, involves embodied, modality-specific, i.e., non-abstract representations. In coming to understand a particular phrase or sentence, a prior experience has to be simulated or reenacted. The representation thus activated is embodied (modality-specific) as sensorimotor regions of the brain are thereby recruited. In this paper, we will first discuss the notion of representation, clarify what it takes for a representation to be embodied or abstract, and distinguish between conceptual and (other) linguistic representations. We will then put forward a characterization of cognitive control and examine its representational infrastructure. The remainder of the paper will be devoted to arguing that language augments cognitive control. To that end, we will draw on two lines of research, which investigate how language augments cognitive control: (i) research on the availability of linguistic labels and (ii) research on the active usage of a linguistic code, specifically, in inner speech. Eventually, we will argue that the cognition-enhancing capacity of language can be explained once we assume that it provides us with (a) abstract, non-embodied representations and with (b) abstract, sparse linguistic representations that may serve as easy-to-manipulate placeholders for fully embodied or otherwise more detailed representations.

Structured sequence processing and combinatorial binding: neurobiologically and computationally informed hypotheses

Understanding how the brain forms representations of structured information distributed in time is a challenging endeavour for the neuroscientific community, requiring computationally and neurobiologically informed approaches. The neural mechanisms for segmenting continuous streams of sensory input and establishing representations of dependencies remain largely unknown, as do the transformations and computations occurring between the brain regions involved in these aspects of sequence processing. We propose a blueprint for a neurobiologically informed and informing computational model of sequence processing (entitled: Vector-symbolic Sequencing of Binding INstantiating Dependencies, or VS-BIND). This model is designed to support the transformation of serially ordered elements in sensory sequences into structured representations of bound dependencies, readily operates on multiple timescales, and encodes or decodes sequences with respect to chunked items wherever dependencies occur in time. The model integrates established vector symbolic additive and conjunctive binding operators with neurobiologically plausible oscillatory dynamics, and is compatible with modern spiking neural network simulation methods. We show that the model is capable of simulating previous findings from structured sequence processing tasks that engage fronto-temporal regions, specifying mechanistic roles for regions such as prefrontal areas 44/45 and the frontal operculum during interactions with sensory representations in temporal cortex. Finally, we are able to make predictions based on the configuration of the model alone that underscore the importance of serial position information, which requires input from time-sensitive cells, known to reside in the hippocampus and dorsolateral prefrontal cortex. This article is part of the theme issue 'Towards mechanistic models of meaning composition'.

Cognitive Control Deficits in Alcohol Dependence Are a Trait- and State-Dependent Biomarker: An ERP Study

Alcohol dependence (AD) presents cognitive control deficits. Event-related potential (ERP) P300 reflects cognitive control-related processing. The aim of this study was to investigate whether cognitive control deficits are a trait biomarker or a state biomarker in AD. Participants included 30 AD patients and 30 healthy controls (HCs). All participants were measured with P300 evoked by a three-stimulus auditory oddball paradigm at a normal state (time 1, i.e., just after the last alcohol intake) and abstinence (time 2, i.e., just after a 4-week abstinence). The results showed that for P3a and P3b amplitude, the interaction effect for group × time point was significant, the simple effect for group at time 1 level and time 2 level was significant, and the simple effect for time point at AD group level was significant; however, the simple effect for time point at HC group level was not significant. Above results indicated that compared to HCs, AD patients present reductions of P3a/3b amplitude, and after 4-week alcohol abstinence, although P3a/3b amplitudes were improved, they were still lower than those of HCs. For P3a and P3b latencies, no significant differences were observed. These findings conclude that AD patients present cognitive control deficits that are reflected by P3a/3b and that cognitive control deficits in AD are trait- and state-dependent. The implication of these findings is helpful to understand the psychological and neural processes for AD, and these findings suggest that improving the cognitive control function may impact the treatment effect for AD.

Why does the mind wander?

I seek an explanation for the etiology and the function of mind wandering episodes. My proposal-which I call the cognitive control proposal-is that mind wandering is a form of non-conscious guidance due to cognitive control. When the agent's current goal is deemed insufficiently rewarding, the cognitive control system initiates a search for a new, more rewarding goal. This search is the process of unintentional mind wandering. After developing the proposal, and relating it to the literature on mind wandering and on cognitive control, I discuss explanations the proposal affords, testable predictions the proposal makes, and philosophical implications the proposal has.

Gated spiking neural network using Iterative Free-Energy Optimization and rank-order coding for structure learning in memory sequences (INFERNO GATE)

We present a framework based on iterative free-energy optimization with spiking neural networks for modeling the fronto-striatal system (PFC-BG) for the generation and recall of audio memory sequences. In line with neuroimaging studies carried out in the PFC, we propose a genuine coding strategy using the gain-modulation mechanism to represent abstract sequences based solely on the rank and location of items within them. Based on this mechanism, we show that we can construct a repertoire of neurons sensitive to the temporal structure in sequences from which we can represent any novel sequences. Free-energy optimization is then used to explore and to retrieve the missing indices of the items in the correct order for executive control and compositionality. We show that the gain-modulation mechanism permits the network to be robust to variabilities and to have long-term dependencies as it implements a gated recurrent neural network. This model, called Inferno Gate, is an extension of the neural architecture Inferno standing for Iterative Free-Energy Optimization of Recurrent Neural Networks with Gating or Gain-modulation. In experiments performed with an audio database of ten thousand MFCC vectors, Inferno Gate is capable of encoding efficiently and retrieving chunks of fifty items length. We then discuss the potential of our network to model the features of working memory in the PFC-BG loop for structural learning, goal-direction and hierarchical reinforcement learning.

A dual architecture for the cognitive control of language: Evidence from functional imaging and language production

The relation between language processing and the cognitive control of thought and action is a widely debated issue in cognitive neuroscience. While recent research suggests a modular separation between a 'language system' for meaningful linguistic processing and a 'multiple-demand system' for cognitive control, other findings point to more integrated perspectives in which controlled language processing emerges from a division of labor between (parts of) the language system and (parts of) the multiple-demand system. We test here a dual approach to the cognitive control of language predicated on the notion of cognitive control as the combined contribution of a semantic control network (SCN) and a working memory network (WMN) supporting top-down manipulation of (lexico-)semantic information and the monitoring of information in verbal working memory, respectively. We reveal these networks in a large-scale coordinate-based meta-analysis contrasting functional imaging studies of verbal working memory vs. active judgments on (lexico-)semantic information and show the extent of their overlap with the multiple-demand system and the language system. Testing these networks' involvement in a functional imaging study of object naming and verb generation, we then show that SCN specializes in top-down retrieval and selection of (lexico-)semantic representations amongst competing alternatives, while WMN intervenes at a more general level of control modulated in part by the amount of competing responses available for selection. These results have implications in conceptualizing the neurocognitive architecture of language and cognitive control.