"Not" in the brain and behavior

Cas W. Coopmans,Anna Mai,Andrea E. Martin

DOI: https://doi.org/10.1371/journal.pbio.3002656

IF: 9.8

2024-06-01

PLoS Biology

Abstract:Negation is key for cognition but has no physical basis, raising questions about its neural origins. A new study in PLOS Biology on the negation of scalar adjectives shows that negation acts in part by altering the response to the adjective it negates.

biochemistry & molecular biology,biology

What problem does this paper attempt to address?

Negation mitigates rather than inverts the neural representations of adjectives

Arianna Zuanazzi,Pablo Ripollés,Wy Ming Lin,Laura Gwilliams,Jean-Rémi King,David Poeppel

DOI: https://doi.org/10.1371/journal.pbio.3002622

IF: 9.8

2024-05-31

PLoS Biology

Abstract:Combinatoric linguistic operations underpin human language processes, but how meaning is composed and refined in the mind of the reader is not well understood. We address this puzzle by exploiting the ubiquitous function of negation. We track the online effects of negation ("not") and intensifiers ("really") on the representation of scalar adjectives (e.g., "good") in parametrically designed behavioral and neurophysiological (MEG) experiments. The behavioral data show that participants first interpret negated adjectives as affirmative and later modify their interpretation towards, but never exactly as, the opposite meaning. Decoding analyses of neural activity further reveal significant above chance decoding accuracy for negated adjectives within 600 ms from adjective onset, suggesting that negation does not invert the representation of adjectives (i.e., "not bad" represented as "good"); furthermore, decoding accuracy for negated adjectives is found to be significantly lower than that for affirmative adjectives. Overall, these results suggest that negation mitigates rather than inverts the neural representations of adjectives. This putative suppression mechanism of negation is supported by increased synchronization of beta-band neural activity in sensorimotor areas. The analysis of negation provides a steppingstone to understand how the human brain represents changes of meaning over time.

biochemistry & molecular biology,biology
Negation in the brain: modulating action representations

Marco Tettamanti,Rosa Manenti,Pasquale A Della Rosa,Andrea Falini,Daniela Perani,Stefano F Cappa,Andrea Moro

DOI: https://doi.org/10.1016/j.neuroimage.2008.08.004

IF: 5.7

2008-11-01

NeuroImage

Abstract:Sentential negation is a universal syntactic feature of human languages that reverses the truth value expressed by a sentence. An intriguing question concerns what brain mechanisms underlie our ability to represent and understand the meaning of negative sentences. We approach this issue by investigating action-related language processing and the associated neural representations. Using functional magnetic resonance imaging we measured brain activity in 18 healthy subjects during passive listening of sentences characterized by a factorial combination of polarity (affirmative vs. negative) and concreteness (action-related vs. abstract). Negation deactivated cortical areas and the left pallidum. Compared to abstract sentences, action-related sentences activated the left-hemispheric action-representation system. Crucially, the polarity by concreteness interactions showed that the activity within the action-representation system was specifically reduced for negative action-related vs. affirmative action-related sentences (compared to abstract sentences). Accordingly, functional integration within this system as measured by Dynamic Causal Modeling was specifically weaker for negative action-related than for affirmative action-related sentences. This modulation of action representations indicates that sentential negation transiently reduces the access to mental representations of the negated information.
The Neurobiology of Lexical and Sentential Negation

Liuba Papeo,Manuel de Vega

DOI: https://doi.org/10.1093/oxfordhb/9780198830528.013.44

2020-03-25

Abstract:Is negation inherently harder than affirmation? How does negation change the neural representation of a word meaning? Recent studies considering pragmatic aspects of language, largely neglected in early studies on negation, have challenged the view that negation involves an extra processing stage to modify, or recode, an affirmative representation initially activated. In consequence, researchers have sought to determine the earliest effects of negation on the neural activity associated with language understanding. Behavioral, neuroimaging, and electrophysiological research have highlighted three circumstances. First, negation changes the neural representation of a word meaning from the earliest stages of word processing. Second, negation reduces or abolishes the activity in brain areas representing the word in the scope of negation. Third, the brain network for response inhibition is recruited during processing negation. Composing these results into a unitary framework, a neurobiological model of negation is proposed that integrates the neurophysiological mechanisms of lexical-semantic processing and inhibition .
The Default Computation of Negated Meanings

Liuba Papeo,Jean-Rémy Hochmann,Lorella Battelli

DOI: https://doi.org/10.1162/jocn_a_01016

2016-12-01

Journal of Cognitive Neuroscience

Abstract:Negation is a fundamental component of human reasoning and language. Yet, current neurocognitive models, conceived to account for the cortical representation of meanings (e.g., writing), hardly accommodate the representation of negated meanings (not writing). One main hypothesis, known as the two-step model, proposes that, for negated meanings, the corresponding positive representation is first fully activated and then modified to reflect negation. Recast in neurobiological terms, this model predicts that, in the initial stage of semantic processing, the neural representation of a stimulus' meaning is indistinguishable from the neural representation of that meaning following negation. Although previous work has shown that pragmatic and task manipulations can favor or hinder a two-step processing, we just do not know how the brain processes an utterance as simple as “I am not writing.” We implemented two methodologies based on chronometric TMS to measure motor excitability (Experiment 1) and inhibition (Experiment 2) as physiological markers of semantic access to action-related meanings. We used elementary sentences (Adverb + Verb) and a passive reading task. For the first time, we defined action word-related motor activity in terms of increased excitability and concurrently reduced inhibition. Moreover, we showed that this pattern changes already in the earliest stage of semantic processing, when action meanings were negated. Negation modifies the neural representation of the argument in its scope, as soon as semantic effects are observed in the brain.

psychology, experimental,neurosciences
Brain Inhibitory Mechanisms Are Involved in the Processing of Sentential Negation, Regardless of Its Content. Evidence From EEG Theta and Beta Rhythms

David Beltrán,Yurena Morera,Enrique García-Marco,Manuel de Vega

DOI: https://doi.org/10.3389/fpsyg.2019.01782

IF: 3.8

2019-08-08

Frontiers in Psychology

Abstract:The two-step process account of negation understanding posits an initial representation of the negated events, followed by a representation of the actual state of events. On the other hand, behavioral and neurophysiological studies provided evidence that linguistic negation suppresses or reduces the activation of the negated events, contributing to shift attention to the actual state of events. However, the specific mechanism of this suppression is poorly known. Recently, based on the brain organization principle of neural reuse (<a href="#B1">Anderson, 2010</a>), it has been proposed that understanding linguistic negation partially relies upon the neurophysiological mechanisms of response inhibition. Specifically, it was reported that negated action-related sentences modulate EEG signatures of response inhibition (<a href="#B15">de Vega et al., 2016</a>; <a href="#B7">Beltrán et al., 2018</a>). In the current EEG study, we ponder whether the reusing of response inhibition processes by negation is constrained to action-related contents or consists of a more general-purpose mechanism. To this end, we employed the same dual-task paradigm as in our prior study—a Go/NoGo task embedded into a sentence comprehension task—but this time including both action and non-action sentences. The results confirmed that the increase of theta power elicited by NoGo trials was modulated by negative sentences, compared to their affirmative counterparts, and this polarity effect was statistically similar for both action- and non-action-related sentences. Thus, a general-purpose inhibitory control mechanism, rather than one specific for action language, is likely operating in the comprehension of sentential negation to produce the transition between alternative representations.

psychology, multidisciplinary
Inhibitory Mechanisms in the Processing of Negations: A Neural Reuse Hypothesis

David Beltrán,Bo Liu,Manuel de Vega

DOI: https://doi.org/10.1007/s10936-021-09796-x

2021-08-12

Journal of Psycholinguistic Research

Abstract:Abstract Negation is known to have inhibitory consequences for the information under its scope. However, how it produces such effects remains poorly understood. Recently, it has been proposed that negation processing might be implemented at the neural level by the recruitment of inhibitory and cognitive control mechanisms. On this line, this manuscript offers the hypothesis that negation reuses general-domain mechanisms that subserve inhibition in other non-linguistic cognitive functions. The first two sections describe the inhibitory effects of negation on conceptual representations and its embodied effects , as well as the theoretical foundations for the reuse hypothesis. The next section describes the neurophysiological evidence that linguistic negation interacts with response inhibition, along with the suggestion that both functions share inhibitory mechanisms. Finally, the manuscript concludes that the functional relation between negation and inhibition observed at the mechanistic level could be easily integrated with predominant cognitive models of negation processing.

linguistics,psychology, experimental
Negation as a means for emotion regulation? Startle reflex modulation during processing of negated emotional words

Cornelia Herbert,Roland Deutsch,Stefan Sütterlin,Andrea Kübler,Paul Pauli

DOI: https://doi.org/10.3758/s13415-011-0026-1

Abstract:This study investigated startle reflex modulation in 33 healthy student participants during the processing of negated emotional items. To build upon previous research, our particular interest was to find out whether processing of negated emotional items modulates emotional responding in line with the logical meaning of the negated expression, or instead leads to paradox emotional effects that point in the direction opposite the one logically implied by the negation. Startle reflex modulation was assessed during silent reading of pleasant and unpleasant nouns. The nouns were either paired with the possessive pronoun my or with the negation word no. The startle eyeblink amplitude was enhanced during processing of the unpleasant pronoun-noun phrases and attenuated during processing of the pleasant phrases. Negation attenuated the startle eyeblink for negated unpleasant nouns and enhanced it for negated pleasant nouns. In line with this finding, negation decreased arousal ratings for unpleasant nouns and reversed the valence ratings for pleasant nouns. Our results are the first to show an effect of negation on both peripheral physiological and subjective indices of affective responding. Our results suggest that negation may be an effective strategy for spontaneous down-regulation of emotional responses to unpleasant, but not to pleasant, stimuli.
Are Scalar Implicatures Automatically Processed and Different for Each Individual? A Mismatch Negativity (MMN) Study.

Ming Zhao,Tao Liu,Gang Chen,Feiyan Chen

DOI: https://doi.org/10.1016/j.brainres.2014.11.049

IF: 3.61

2014-01-01

Brain Research

Abstract:Scalar implicatures are ordinarily activated in human communication when the speaker uses a weak expression (e.g., some) from a set of stronger alternatives (e.g., many, all). It has been debated whether scalar inferences are generated by default. To clarify this issue and examine whether individual pragmatic ability will affect the mechanism of scalar inference processing, we performed experiment with an MMN paradigm to capture the neurophysiological indicators of automatic processing of spoken sentences and divided participants into high and low pragmatic ability groups. Experimental results showed that compared with the condition that an informative sentence ("Some animals have tails") is the deviant stimuli, when an underinformative sentence ("Some tigers have tails") is the deviant stimuli, the high pragmatic ability group induced mismatch negativity (MMN) and sustained negativity, while the low pragmatic ability group had no ERP effects. These results indicated that at least some people can automatically activate the scalar implicatures when encountering scalar trigger words, even in the inattentive status.
Effects of Semantic and Pragmatic Factors on Preschool Children’s Negation-Triggered Inferences on Possible Alternatives

Xiaowen Zhang,Peng Zhou

DOI: https://doi.org/10.1016/j.jecp.2024.106057

IF: 2.547

2024-01-01

Journal of Experimental Child Psychology

Abstract:Negation-triggered inferences are universal across human languages. Hearing “This is not X” should logically lead to the inference that all elements other than X constitute possible alternatives. However, not all logically possible alternatives are equally accessible in the real world. To qualify as a plausible alternative, it must share with the negated element as many similarities as possible, and the most plausible one is often from the same taxonomic category as the negated element. The current article reports on two experiments that investigated the development of preschool children’s ability to infer plausible alternatives triggered by negation. Experiment 1 showed that in a context where children were required to determine the most plausible alternative to the negated element, the 3-, 4- and 5-year-olds, exhibited a robust preference for the taxonomic associates. Experiment 2 further demonstrated that the 3-, 4- and 5-year-olds considered all the complement set members as equally possible alternatives in a context where they were not explicitly required to evaluate the plausibility of different candidates. Taken together, our findings reveal interesting developmental continuity in preschool children’s ability to make inferences about plausible alternatives triggered by negation. We discuss the potential semantic and pragmatic factors that contribute to children’s emerging awareness of typical alternatives triggered by negative expressions.
The neural inhibition network is causally involved in the disembodiment effect of linguistic negation

Francesca Vitale,Ilaria Monti,Iván Padrón,Alessio Avenanti,Manuel de Vega

DOI: https://doi.org/10.1016/j.cortex.2021.11.015

IF: 4.644

Cortex

Abstract:Negation applied to action contexts reduces the activation of the motor system. According to the Reusing Inhibition for Negation (RIN) hypothesis, such "disembodiment" effect occurs because understanding negations engages the reuse of inhibitory control mechanisms. Here, we investigated whether the right inferior frontal gyrus (rIFG) - a key area of the inhibitory control system - contributes to primary motor cortex (M1) processing of negated action-sentences. Using a perturb-and-measure paradigm, we applied off-line low-frequency repetitive TMS (rTMS) over the rIFG, before performing a reading task involving action and attentional sentences presented in both affirmative or negative form. During the reading task, motor excitability was assessed by recording motor-evoked potentials (MEPs) induced by single-pulse TMS (spTMS) over the left M1, at two loci in the sentence: the verb or the object. Results show that after sham stimulation (baseline), motor excitability measured on the verb, was reduced for negative, compared to affirmative action sentences. Crucially, neuromodulation of rIFG suppressed this inhibitory effect of negation, since motor excitability was equaled for negative and affirmative action sentences. As expected, no effect of negation was observed for attentional sentences or when the pulse was delivered over the object. Our study confirms that understanding negative action sentences inhibits M1. This effect took place at an early stage of semantic processing (i.e., while processing the verb in our task), and faded at a later time-point. Critically, by highlighting a causal role of rIFG in this motor inhibition, we provide direct neurophysiological support to the RIN hypothesis.
Mental simulation of the factual and the illusory in negation processing: evidence from anticipatory eye movements on a blank screen

Norbert Vanek,Ana Matić Škorić,Sara Košutar,Štěpán Matějka,Kate Stone

DOI: https://doi.org/10.1038/s41598-024-53353-0

IF: 4.6

2024-02-04

Scientific Reports

Abstract:How do comprehenders process negative statements such as The fish is not jumping out of the water ? Opinions vary. Some argue for two steps, namely that processing starts off with the representation of the positive/illusory [fish jumping out of the water] and then shifts to the (f)actual. To test this idea, we measured fixations on the factual (fish not jumping) versus the illusory (fish jumping) during auditory processing of negation and affirmation. We tested speakers of English (single-cued negation) and Croatian (double-cued negation) and focused on anticipatory fixations in the absence of pictures to indicate the strength of mental simulations. Our findings contribute to negation processing research in two ways. First, dominant anticipatory fixations on the factual suggest a direct rather than a two-step process. Second, time-sensitive insights from two languages call for a finer-grained account of negation processing with negation-specific support of inferences of the factual over the illusory.

multidisciplinary sciences
Presetting an inhibitory state modifies the neural processing of negated action sentences. An ERP study

Bo Liu,Beixian Gu,David Beltrán,Huili Wang,Manuel de Vega

DOI: https://doi.org/10.1016/j.bandc.2020.105598

Abstract:It has been proposed that understanding negated action sentences (You don't cut the bread) uses the neural networks of action inhibition. The evidence comes from studies in which affirmative or negative action language immediately preceded a Go/NoGo task. It was found that negation selectively modulates inhibition-related signatures of NoGo trials, supporting the Reusing Inhibition for Negation (RIN) hypothesis. To further explore this hypothesis, this study tested the reverse effects; namely, how presetting an inhibitory state affects the processing of negated action sentences. To this end, Go/NoGo responses preceded sentence reading and EEG activities were recorded throughout the entire trials. ERP results indicate that the presetting of inhibition by the NoGo cue induced a sustained modulation of waveform for negated action sentences relative to affirmative ones, which began shortly after the negation operator onset and remained beyond the action verb onset. Crucially, the estimated sources of such effect were the right inferior frontal gyrus and the left middle frontal gyrus, both relevant regions in the action inhibition circuitry. These results, complemented by previous findings, support the idea that action inhibition and negated action language share neural mechanisms and influence each other, thus confirming and extending the RIN hypothesis.
Disentangling inhibition and prediction in negation processing

Viviana Haase

DOI: https://doi.org/10.3389/flang.2024.1304613

2024-03-01

Abstract:Negation as a universal feature of human language is used effortlessly in everyday communication. However, experimental research has shown that the comprehension of negated sentences seems to require additional cognitive resources compared to affirmative sentences. Many studies investigating the processing of negation report longer reading and reaction times for negative compared to affirmative sentences and many studies report a Polarity by Truth interaction: false affirmative sentences lead to longer response times and larger N400 event-related potentials (ERPs) than true affirmative sentences, whereas the pattern is reversed for negative sentences where it is the true sentence that elicits longer reaction times and higher N400 ERPs compared to false negative sentences. These interactions have been discussed in the light of lexical associations, predictability, and the need to construct two subsequent mental representations. Furthermore, recent studies have shown that the comprehension of negated sentences seems to make use of neural resources that are typically involved in cognitive control and inhibitory mechanisms. As both processes have been associated with two different and temporally overlapping ERP components (the N400 and the P300), we focus on studies with high temporal resolution. We discuss linguistic aspects of negation, such as semantic similarity and contextual invariance of negation. We furthermore discuss the role of the verb as well as the position of the negative marker with respect to the verb, and their respective relevance for predictive and inhibitory mechanisms in negated sentences.
Embodied negation and levels of concreteness: A TMS Study on German and Italian language processing

Giorgio Papitto,Luisa Lugli,Anna M. Borghi,Antonello Pellicano,Ferdinand Binkofski

DOI: https://doi.org/10.1016/j.brainres.2021.147523

2021-06-08

Abstract:According to the embodied cognition perspective, linguistic negation may block the motor simulations induced by language processing. Transcranial magnetic stimulation (TMS) was applied to the left primary motor cortex (hand area) of monolingual Italian and German healthy participants during a rapid serial visual presentation of sentences from their own language. In these languages, the negative particle is located at the beginning and at the end of the sentence, respectively. The study investigated whether the interruption of the motor simulation processes, accounted for by reduced motor evoked potentials (MEPs), takes place similarly in two languages differing on the position of the negative marker. Different levels of sentence concreteness were also manipulated to investigate if negation exerts generalized effects or if it is affected by the semantic features of the sentence. Our findings indicate that negation acts as a block on motor representations, but independently from the language and words concreteness level.

Neurons and Cognition
What is in a word? No versus Yes differentially engage the lateral orbitofrontal cortex.

Nelly Alia-Klein,Rita Z. Goldstein,Dardo Tomasi,Lei Zhang,Stephanie Fagin-Jones,Frank Telang,Gene-Jack Wang,Joanna S. Fowler,Nora D. Volkow

DOI: https://doi.org/10.1037/1528-3542.7.3.649

IF: 5.564

2007-08-01

Emotion

Abstract:The words "No" and "Yes" are involved in conditioning to prohibit or encourage behavior, respectively. The authors, therefore, hypothesized that these words would be attributed to endogenous valence, activating neuronal circuits involved with valence and emotional control. Functional MRI (fMRI) at 4 Tesla was used to record regional brain activity while participants were exposed to emphatic vocalizations of the words. Results showed that No and Yes were associated with opposite brain-behavior responses; while No was negatively valenced, produced slower response times, and evoked a negative signal in the right lateral orbitofrontal cortex (OFC), Yes was positively valenced, produced faster response times, and evoked a positive signal in a contiguous region of the OFC. Attribution of negative valence to No and trait anger control were associated with increased responsivity of the OFC to No. Inasmuch as sensitivity to the prohibitive command No develops during childhood through interaction with primary caregivers as the first social objects, our findings may implicate the lateral OFC in the neurobiology of emotion regulation and subsequent social development.

psychology, experimental
Looks at what isn't there: eye movements on a blank screen when processing negation in a first and a second language

Norbert Vanek,Ana Matić Škorić,Sara Košutar,Štěpán Matějka,Kate Stone

DOI: https://doi.org/10.3389/fnhum.2024.1457038

IF: 3.473

2024-10-10

Frontiers in Human Neuroscience

Abstract:Is negation more difficult to process than affirmation? If it is, does processing negation in a second language (L2) compound the difficulty compared to the first language (L1)? This article addresses the issues of difficulties in processing different types of negation in the L1 and L2 by looking at the differences in the ways in which comprehenders anticipate upcoming visual information during sentence processing. Using a blank-screen paradigm, we recorded eye fixations of Croatian native speakers and Croatian learners of English while they were anticipating matching or mismatching pictures to sentences with various types of negation in L1 (Croatian) and L2 (English). Using a between-group design, we manipulated sentence polarity (affirmative vs. negative), negation type (sentential vs. negative quantifier) within both L1 Croatian and L2 English so that we could observe potential anticipation effects varying as a function of the two predictors. In line with previous studies, affirmation in the L1 was easier to process than negation, and participants were able to anticipate sentence-picture matches in both the L1 and the L2 group. In contrast with our prediction, anticipatory looks did not significantly vary across negation types in Croatian based on the number of structural cues. In L2 English, learners exhibited prediction ability across negation types. These findings go against the view that comprehension in L2 comes with a reduced ability to generate expectations, and they highlight the robustness of mental simulations in both L1 and L2 negation processing.

psychology,neurosciences
Negation in Exclamatives

Gaetano Fiorin,D. Delfitto

DOI: https://doi.org/10.1111/STUL.12024

2014-12-01

Abstract:The goal of this article is to provide a principled account of the semantic contribution of the so-called “expletive” negation, as it occurs in negative exclamatives. This type of negation differs from canonical negation in that it does not contribute its canonical meaning as a propositional operator, occurs higher in the syntactic structure of the clause, does not license negative polarity items, and expresses a bias towards the opposite polarity. In this article, we show that these properties are also found in negative rhetorical questions. We propose that exclamatives, as well as rhetorical questions, denote sets of propositions structured into Boolean algebras, with the negation imposing an order of informativity upon such structures.

Mathematics,Linguistics
The role of motor inhibition in implicit negation processing: two Go/No-Go behavioral studies

Martina Montalti,Marta Calbi,Maria Alessandra Umiltà,Vittorio Gallese,Valentina Cuccio

DOI: https://doi.org/10.1007/s00426-024-01941-0

2024-03-15

Psychological Research

Abstract:Several studies demonstrated that explicit forms of negation processing (e.g., "I don't know") recruits motor inhibitory mechanisms. However, whether this is also true for implicit negation, in which the negative meaning is implicated but not explicitly lexicalized in the sentence (e.g., "I ignore"), has never been studied before. Two Go/No-Go studies, which differed only for the time-windows to respond to the Go stimulus, were carried out. In each, participants ( N = 86 in experiment 1; N = 87 in experiment 2) respond to coloured circle while reading task-irrelevant affirmative, explicit negative and implicit negative sentences. We aimed to investigate whether: (i) the processing of implicit negations recruits inhibitory mechanisms; (ii) these inhibitory resources are differently modulated by implicit and explicit negations. Results show that implicit negative sentences recruit the inhibitory resources more strongly when compared to explicit ones, probably due to their inferential nature, likely requiring deeper processing of the negative meaning. Implicit and inferential meaning (i.e., pragmatic information) are grounded too in the same mechanisms that integrate action with perception. Such findings provide further evidence to the embodied account of language, showing that even abstract aspects, like implicit negation, are grounded in the sensory-motor system, by means of functional link between language and motor activity.

psychology, experimental
Robots Learning to Say `No': Prohibition and Rejective Mechanisms in Acquisition of Linguistic Negation

Frank Förster,Joe Saunders,Hagen Lehmann,Chrystopher L. Nehaniv

DOI: https://doi.org/10.1145/3359618

2018-10-28

Abstract:`No' belongs to the first ten words used by children and embodies the first active form of linguistic negation. Despite its early occurrence the details of its acquisition process remain largely unknown. The circumstance that `no' cannot be construed as a label for perceptible objects or events puts it outside of the scope of most modern accounts of language acquisition. Moreover, most symbol grounding architectures will struggle to ground the word due to its non-referential character. In an experimental study involving the child-like humanoid robot iCub that was designed to illuminate the acquisition process of negation words, the robot is deployed in several rounds of speech-wise unconstrained interaction with naïve participants acting as its language teachers. The results corroborate the hypothesis that affect or volition plays a pivotal role in the socially distributed acquisition process. Negation words are prosodically salient within prohibitive utterances and negative intent interpretations such that they can be easily isolated from the teacher's speech signal. These words subsequently may be grounded in negative affective states. However, observations of the nature of prohibitive acts and the temporal relationships between its linguistic and extra-linguistic components raise serious questions over the suitability of Hebbian-type algorithms for language grounding.

Computation and Language,Artificial Intelligence,Robotics
NevIR: Negation in Neural Information Retrieval

Orion Weller,Dawn Lawrie,Benjamin Van Durme

DOI: https://doi.org/10.48550/arXiv.2305.07614

2024-02-27

Abstract:Negation is a common everyday phenomena and has been a consistent area of weakness for language models (LMs). Although the Information Retrieval (IR) community has adopted LMs as the backbone of modern IR architectures, there has been little to no research in understanding how negation impacts neural IR. We therefore construct a straightforward benchmark on this theme: asking IR models to rank two documents that differ only by negation. We show that the results vary widely according to the type of IR architecture: cross-encoders perform best, followed by late-interaction models, and in last place are bi-encoder and sparse neural architectures. We find that most information retrieval models (including SOTA ones) do not consider negation, performing the same or worse than a random ranking. We show that although the obvious approach of continued fine-tuning on a dataset of contrastive documents containing negations increases performance (as does model size), there is still a large gap between machine and human performance.

Information Retrieval,Computation and Language

"Not" in the brain and behavior

Negation mitigates rather than inverts the neural representations of adjectives

Negation in the brain: modulating action representations

The Neurobiology of Lexical and Sentential Negation

The Default Computation of Negated Meanings

Brain Inhibitory Mechanisms Are Involved in the Processing of Sentential Negation, Regardless of Its Content. Evidence From EEG Theta and Beta Rhythms

Inhibitory Mechanisms in the Processing of Negations: A Neural Reuse Hypothesis

Negation as a means for emotion regulation? Startle reflex modulation during processing of negated emotional words

Are Scalar Implicatures Automatically Processed and Different for Each Individual? A Mismatch Negativity (MMN) Study.

Effects of Semantic and Pragmatic Factors on Preschool Children’s Negation-Triggered Inferences on Possible Alternatives

The neural inhibition network is causally involved in the disembodiment effect of linguistic negation

Mental simulation of the factual and the illusory in negation processing: evidence from anticipatory eye movements on a blank screen

Presetting an inhibitory state modifies the neural processing of negated action sentences. An ERP study

Disentangling inhibition and prediction in negation processing

Embodied negation and levels of concreteness: A TMS Study on German and Italian language processing

What is in a word? No versus Yes differentially engage the lateral orbitofrontal cortex.

Looks at what isn't there: eye movements on a blank screen when processing negation in a first and a second language

Negation in Exclamatives

The role of motor inhibition in implicit negation processing: two Go/No-Go behavioral studies

Robots Learning to Say `No': Prohibition and Rejective Mechanisms in Acquisition of Linguistic Negation

NevIR: Negation in Neural Information Retrieval