Publications

Abstract

According to the language marker hypothesis language has provided homo sapiens with a rich symbolic system that plays a central role in interpreting signals delivered by our sensory apparatus, in shaping action goals, and in creating a powerful tool for reasoning and inferencing. This view provides an important correction on embodied accounts of language that reduce language to action, perception, emotion and mental simulation. The presence of a language system has, however, also important consequences for perception, action, emotion, and memory. Language stamps signals from perception, action, and emotional systems with rich cognitive markers that transform the role of these signals in the overall cognitive architecture of the human mind. This view does not deny that language is implemented by means of universal principles of neural organization. However, language creates the possibility to generate rich internal models of the world that are shaped and made accessible by the characteristics of a language system. This makes us less dependent on direct action-perception couplings and might even sometimes go at the expense of the veridicality of perception. In cognitive (neuro)science the pendulum has swung from language as the key to understand the organization of the human mind to the perspective that it is a byproduct of perception and action. It is time that it partly swings back again.

Abstract

EEG and eye-tracking provide complementary information when investigating language comprehension. Evidence that speech processing may be facilitated by speech prediction comes from the observation that a listener's eye gaze moves towards a referent before it is mentioned if the remainder of the spoken sentence is predictable. However, changes to the trajectory of anticipatory fixations could result from a change in prediction or an attention shift. Conversely, N400 amplitudes and concurrent spectral power provide information about the ease of word processing the moment the word is perceived. In a proof-of-principle investigation, we combined EEG and eye-tracking to study linguistic prediction in naturalistic, virtual environments. We observed increased processing, reflected in theta band power, either during verb processing - when the verb was predictive of the noun - or during noun processing - when the verb was not predictive of the noun. Alpha power was higher in response to the predictive verb and unpredictable nouns. We replicated typical effects of noun congruence but not predictability on the N400 in response to the noun. Thus, the rich visual context that accompanied speech in virtual reality influenced language processing compared to previous reports, where the visual context may have facilitated processing of unpredictable nouns. Finally, anticipatory fixations were predictive of spectral power during noun processing and the length of time fixating the target could be predicted by spectral power at verb onset, conditional on the object having been fixated. Overall, we show that combining EEG and eye-tracking provides a promising new method to answer novel research questions about the prediction of upcoming linguistic input, for example, regarding the role of extralinguistic cues in prediction during language comprehension.

Abstract

During listening, brain activity tracks the rhythmic structures of speech signals. Here, we directly dissociated the contribution of neural envelope tracking in the processing of speech acoustic cues from that related to linguistic processing. We examined the neural changes associated with the comprehension of Noise-Vocoded (NV) speech using magnetoencephalography (MEG). Participants listened to NV sentences in a 3-phase training paradigm: (1) pre-training, where NV stimuli were barely comprehended, (2) training with exposure of the original clear version of speech stimulus, and (3) post-training, where the same stimuli gained intelligibility from the training phase. Using this paradigm, we tested if the neural responses of a speech signal was modulated by its intelligibility without any change in its acoustic structure. To test the influence of spectral degradation on neural envelope tracking independently of training, participants listened to two types of NV sentences (4-band and 2-band NV speech), but were only trained to understand 4-band NV speech. Significant changes in neural tracking were observed in the delta range in relation to the acoustic degradation of speech. However, we failed to find a direct effect of intelligibility on the neural tracking of speech envelope in both theta and delta ranges, in both auditory regions-of-interest and whole-brain sensor-space analyses. This suggests that acoustics greatly influence the neural tracking response to speech envelope, and that caution needs to be taken when choosing the control signals for speech-brain tracking analyses, considering that a slight change in acoustic parameters can have strong effects on the neural tracking response.

Abstract

Affective behaviors enable social robots to not only establish better connections with humans but also serve as a tool for the robots to express their internal states. It has been well established that emotions are important to signal understanding in Human-Robot Interaction (HRI). This work aims to harness the power of Large Language Models (LLM) and proposes an approach to control the affective behavior of robots. By interpreting emotion appraisal as an Emotion Recognition in Conversation (ERC) tasks, we used GPT-3.5 to predict the emotion of a robot’s turn in real-time, using the dialogue history of the ongoing conversation. The robot signaled the predicted emotion using facial expressions. The model was evaluated in a within-subjects user study (N = 47) where the model-driven emotion generation was compared against conditions where the robot did not display any emotions and where it displayed incongruent emotions. The participants interacted with the robot by playing a card sorting game that was specifically designed to evoke emotions. The results indicated that the emotions were reliably generated by the LLM and the participants were able to perceive the robot’s emotions. It was found that the robot expressing congruent model-driven facial emotion expressions were perceived to be significantly more human-like, emotionally appropriate, and elicit a more positive impression. Participants also scored significantly better in the card sorting game when the robot displayed congruent facial expressions. From a technical perspective, the study shows that LLMs can be used to control the affective behavior of robots reliably in real-time. Additionally, our results could be used in devising novel human-robot interactions, making robots more effective in roles where emotional interaction is important, such as therapy, companionship, or customer service.

Abstract

Neuron models with explicit dendritic dynamics have shed light on mechanisms for coincidence detection, pathway selection and temporal filtering. However, it is still unclear which morphological and physiological features are required to capture these phenomena. In this work, we introduce the Tripod neuron model and propose a minimal structural reduction of the dendritic tree that is able to reproduce these computations. The Tripod is a three-compartment model consisting of two segregated passive dendrites and a somatic compartment modelled as an adaptive, exponential integrate-and-fire neuron. It incorporates dendritic geometry, membrane physiology and receptor dynamics as measured in human pyramidal cells. We characterize the response of the Tripod to glutamatergic and GABAergic inputs and identify parameters that support supra-linear integration, coincidence-detection and pathway-specific gating through shunting inhibition. Following NMDA spikes, the Tripod neuron generates plateau potentials whose duration depends on the dendritic length and the strength of synaptic input. When fitted with distal compartments, the Tripod encodes previous activity into a dendritic depolarized state. This dendritic memory allows the neuron to perform temporal binding, and we show that it solves transition and sequence detection tasks on which a single-compartment model fails. Thus, the Tripod can account for dendritic computations previously explained only with more detailed neuron models or neural networks. Due to its simplicity, the Tripod neuron can be used efficiently in simulations of larger cortical circuits.

Abstract

The N400 is an endogenous event-related brain potential (ERP) that is sensitive to semantic processes during language comprehension. The general question we address in this paper is which aspects of the comprehension process are manifest in the N400. The focus is on the sensitivity of the N400 to the automatic process of lexical access, or to the controlled process of lexical integration. The former process is the reflex-like and effortless behavior of computing a form representation of the linguistic signal, and of mapping this representation onto corresponding entries in the mental lexicon. The latter process concerns the integration of a spoken or written word into a higher-order meaning representation of the context within which it occurs. ERPs and reaction times (RTs) were acquired to target words preceded by semantically related and unrelated prime words. The semantic relationship between a prime and its target has been shown to modulate the amplitude of the N400 to the target. This modulation can arise from lexical access processes, reflecting the automatic spread of activation between words related in meaning in the mental lexicon. Alternatively, the N400 effect can arise from lexical integration processes, reflecting the relative ease of meaning integration between the prime and the target. To assess the impact of automatic lexical access processes on the N400, we compared the effect of masked and unmasked presentations of a prime on the N400 to a following target. Masking prevents perceptual identification, and as such it is claimed to rule out effects from controlled processes. It therefore enables a stringent test of the possible impact of automatic lexical access processes on the N400. The RT study showed a significant semantic priming effect under both unmasked and masked presentations of the prime. The result for masked priming reflects the effect of automatic spreading of activation during the lexical access process. The ERP study showed a significant N400 effect for the unmasked presentation condition, but no such effect for the masked presentation condition. This indicates that the N400 is not a manifestation of lexical access processes, but reflects aspects of semantic integration processes.

Abstract

Broca′s and Wernicke′s aphasics performed speeded lexical decisions on the third member of auditorily presented triplets consisting of two word primes followed by either a word or a nonword. In three of the four priming conditions, the second prime was a homonym with two unrelated meanings. The relation of the first prime and the target with the two meanings of the homonym was manipulated in the different priming conditions. The two readings of the ambiguous words either shared their grammatical form class (noun-noun ambiguities) or not (noun-verb ambiguities). The silent intervals between the members of the triplets were varied between 100, 500, and 1250 msec. Priming at the shortest interval is mainly attributed to automatic lexical processing, and priming at the longest interval is mainly due to forms of controlled lexical processing. For both Broca′s and Wernicke′s aphasics overall priming effects were obtained at ISIs of 100 and 500 msec, but not at an ISI of 1250 msec. This pattern of results is consistent with the view that both types of aphasics can automatically access the semantic lexicon, but might be impaired in integrating lexical-semantic information into the context. Broca′s aphasics showed a specific impairment in selecting the contextually appropriate reading of noun-verb ambiguities, which is suggested to result from a failure either in the on-line morphological parsing of complex word forms into a stem and an inflection or in the on-line exploitation of the syntactic implications of the inflectional suffix. In a final experiment patients were asked to explicitly judge the semantic relations between a subset of the primes that were used in the lexical decision study. Wernicke′s aphasics performed worse than both Broca′s aphasics and normal controls, indicating a specific impairment for these patients in consciously operating on automatically accessed lexical-semantic information.

Abstract

This paper presents event-related brain potential (ERP) data from an experiment on syntactic processing. Subjects read individual sentences containing one of three different kinds of violations of the syntactic constraints of Dutch. The ERP results provide evidence for M electrophysiological response to syntactic processing that is qualitatively different from established ERP responses to semantic processing. We refer to this electro-physiological manifestation of parsing as the Syntactic Positive Shift (SPS). The SPS was observed in an experiment in which no task demands, other than to read the input, were imposed on the subjects. The pattern of responses to the different kinds of syntactic violations suggests that the SPS indicates the impossibility for the parser to assign the preferred structure to an incoming string of words, irrespective of the specific syntactic nature of this preferred structure. The implications of these findings for further research on parsing are discussed.