Mental Effort and Counterfactuals Modulate Language Understanding in Older Adults

Abstract

Currently, the validity of the proposals of embodied cognition and its relationship with the language

of older adults is unknown. To answer this, we carried out a neurocognitive experiment using EEG,

using linguistic variables. 27 older adults (M = 70.34 years, SD = 4.82, 15 women and 12 men)

read sentences on a computer screen and responded to an activation test.

The results indicate that the linguistic, factual and counterfactual contexts, as well as the corporeal

parameter “mental effort” modulate the understanding of language and participate with greater or

lesser preponderance in different time windows. Furthermore, the counterfactual language seems

to induce emotions that facilitate the processing of high effort while the interaction under mental

effort - counterfactuality represents a negativity that is associated with the N400 component. The

same goes for high-effort factual language.

Keywords: embodiment, aging, imaginability, counterfactuals, ERPs.

Resumen

Actualmente se desconoce la validez de las propuestas de la cognición corpórea y su relación con

el lenguaje de adultos mayores. Para responder a esto realizamos un experimento neurocognitivo

mediante EEG, utilizando variables lingüísticas. 27 adultos mayores (M= 70.34 años, DE= 4.82,

15 mujeres y 12 hombres) leyeron oraciones en la pantalla de un computador y respondieron a una

prueba de activación.

Los resultados indican que los contextos lingüísticos, factual y contrafactual, así como el parámetro

corpóreo “esfuerzo mental” modulan la comprensión del lenguaje y participan con mayor o menor

preponderancia en diferentes ventanas temporales. Además, el lenguaje contrafactual parece

 2

inducir emociones que facilitan el procesamiento del alto esfuerzo mientras que la interacción bajo

esfuerzo mental – contrafactualidad representa una negatividad que se asocia con el componente

N400. Lo mismo ocurre con el lenguaje factual de alto esfuerzo.

Palabras clave: corporeidad, envejecimiento, imaginabilidad, contrafactuales, ERPs.

1. Introduction

Embodied cognition is comprised within the corpus of theories of the cognitive sciences which

propose that many aspects of our cognition are heavily influenced by our body and its interactions

with the environment. This perspective challenges the traditional cognitive science view, which

considers the mind to function mainly on the basis of symbolic manipulation and has yielded the

notion of the human body as computer-esque in its ontogenic development (Araneda-Hinrichs,

2020). Instead, accounts of embodied cognition (alongside embedded, extended, and enactive ones)

argue that it is grounded in multiple ways throughout the body's interactions with the world. While

some prefer to consider it still, due to its recent and brief development, less a well-defined theory

than a research agenda (Shapiro, 2011), due to the fact that even to this day there is no consensus

regarding the implementation of this spectrum of theories of cognition in a research program

throughout which their central assumptions are accounted for; hence, amounting to a paradigmatic

advance remains an ongoing matter of discussion. Notwithstanding, the experimental praxis of

these investigations do bear the commonality of providing the field with new descriptive

information through state-alteration of perception-action systems, of which all corporeal research

assumes the world to be constrained by (Milkowski, 2013; Cowart, 2004; Wilson et al., 2002). The

articulation of said common ground throughout diverse schools and research programs such as

linguistics, psychology, artificial intelligence, robotis, and neurobiology (Cappuccio, 2019; Cox,

2016; Kording et al., 2018; Ionescu & Vasc, 2014) seem to augur a fruitful convergence of visions.
 3

Indeed, it has been widely adopted as a model in linguistic research; as it suggests a linkage

between body and environment to language and our understanding of it -as opposed to viewing

language as an abstract entity that is detached from physical reality- a number of researchers have

incorporated it into psycholinguistic investigations on topics such as the relationship between

physical effort and counterfactual sentences (Urrutia & de Vega, 2012), or the ones between

balance and posture, and their spatial and temporal representations (Miles et al., 2010).

Furthermore, many metaphors and idiomatic expressions have also been shown to reflect physical

and sensory experiences, like "grasping a concept" or "shouldering a burden" (Barsalou et al., 2008;

de Vega, 2021; Fischer & Zwaan, 2008; García & Ibáñez, 2018; Glenberg & Kaschack, 2002).

This approach suggests that language comprehension isn't just an abstract process that occurs in

the brain, but also involves sensorimotor simulations. For example, while reading or hearing the

sentence "John kicked the ball", these words aren’t processed only at a linguistic level, but brain

areas related to physical movement are triggered and engaged, too (as if visualizing or imagining

the action of to kick a ball). Thus, the embodiment perspective posits that meaning is intimately

embedded in the brain's perceptual, motor, and emotional systems (Barsalou et al., 2008; de Vega,

2021; Fischer & Zwaan, 2008; García & Ibáñez, 2018; Glenberg & Kaschack, 2002). From this

perspective, the body and the specific cerebral cortices related to sensation, motor skills, and

emotions have varying degrees of participation in a dimension that ranges from the total

embodiment of semantic processing and representation to full disembodiment. The idea that

sensory and motor experiences form the basis of conceptual knowledge has a long history in

philosophy, psychology, and neuroscience (Allport & Funnell, 1981; Freud, 1891; Locke, 1847;

Wernicke, 1874). In recent years, this proposal has gained new steam under the rubric of

`embodied´ or `situated´ cognition, supported by numerous neuroimaging and behavioral studies

(Salas-Herrera et al., 2020).

 4

Although embodied theories allow a novel approach to the relationship between cognition, body

and context by proposing direct interactions between sensory-motor systems and the representation

(simulation) of meaning, the empirical evidence from behavioral and neurocognitive studies has

tended to concentrate on young adults, leaving aside stages such as older adulthood. In order for

these theories to be considered theories of cognition, and of language in particular, they require

exploration and validation throughout the life cycle (Salas-Herrera, 2015). In this sense, older

population represents a favorable sample to carry out a contrast of possible embodied proposals,

since the declines in sensory-motor systems could directly affect the processing of sentences that

contain challenging semantic content from the point of view of the neural parameters involved

(Gallese & Lakoff, 2005).

Psycholinguistics, oriented towards the study of aging, has been strengthened by research that has

attempted to elucidate the effects of cognitive aging on language comprehension and production

(Véliz, Riffo & Arancibia, 2010). Models emphasizing the functional and computational

architecture of the linguistic system have been successfully tested, particularly at the level of

language comprehension (De Vega, 2008; Fisher & Zwaan, 2008; Glenberg & Gallese, 2012;

Glenberg & Kaschak, 2002). Notions such as working memory, attentional resources, and

metacognition have gained importance (Cheung & Kemper, 1992; Just & Carpenter, 1992; Kemper,

1987; Kemper et al., 2001). It has been determined that there are dimensions of language that are

able to resist the advance of age, such as semantic processing, linguistic comprehension, and

vocabulary (Burke & Shafto, 2008; Glisky, 2007). Psycholinguistics has expanded towards

different theoretical perspectives, drawing on ideas from the neurosciences supported by

experimental behavioral and neurocognitive paradigms (Barsalou, 1999; Gallese & Lakoff, 2005;

Pulvermüller, 1999, 2013; Zwaan, 2004). The discipline is currently being significantly influenced

by theories of embodiment, which propose a different way of looking at language and its
 5

relationship with the body (De Vega, 2008; Fisher & Zwaan, 2008; Glenberg & Gallese, 2012;

Glenberg & Kaschak, 2002).

Studies on cognitive aging and language processes have observed declines in processing speed,

divided attention, executive functions, and phonological retrieval in language production in older

adults (Salthouse, 2000; Greenwood, 2000; Danckert & Craik, 2013; Burke & Shafto, 2008).

However, semantic memory, sustained attention, and recall are generally preserved (Glisky, 2007).

Behavioral research has shown that linguistic comprehension, particularly semantic processing and

vocabulary, tends to be maintained in older adulthood (Burke & Shafto, 2008).

The prevailing view in psycholinguistics suggests that cognitive decline in aging, specifically in

the relationship between language and verbal working memory, is characterized by a decrease in

verbal working memory capacity, which impairs the quality of language processing (Cheung &

Kemper, 1992; Just & Carpenter, 1992; Kemper, 1987; Kemper et al., 2001). An alternative view

challenges this perspective and proposes an integration of working memory models with

consideration of cognitive architecture and linguistic experience (Caplan & Waters, 1999;

MacDonald Christiansen, 2002; Ericsson & Kintsch, 1995). This alternative approach suggests that

individual differences in language comprehension in older adults arise from variations in working

memory efficiency influenced by linguistic experience, rather than total memory capacity (Caplan

& Waters, 1999; MacDonald Christiansen, 2002).

Brain imaging studies have identified brain regions such as the dorsolateral prefrontal cortex,

parietal lobe, premotor cortex, and inferior frontal gyrus as part of the verbal working memory

circuitry (Goldman-Rakic, 1987; D'Esposito & Postle, 2015; Postle, 2015; Fegen, Buchsbaum &

D'Esposito, 2015). The processing speed in cognitive aging is related to various brain structures,

functions, and regions associated with different cognitive domains (Salthouse, 2017; Haász et al.,

2013; Lockhart et al., 2014; Ferreira et al., 2014; Eckert et al., 2010; Papp et al., 2014).
 6

The activation of motor representations would be directly related to action verb processing

(Beauprez & Bidet-Ildei, 2019). Since the motor plans caused by the verbs must be integrated with

the properties of the associated objects, we suppose that aging could affect the estimation of the

direct object, and thus, sentence reading when this implies variable levels of effort, since the

abilities to perform physical effort in older people are impaired compared to young people

(Portegijs et al., 2019). In fact, several studies have shown that older people have difficulties using

imaging and motor simulation (Muffato et al., 2017).

Simulations utilize corporeal or neural parameters, such as direction for reaching and

force/physical effort for grasping an object, imposing a hierarchical structure on the brain (Gallese

& Lakoff, 2005). The term "simulation" refers to mentally and unconsciously simulating the action

or perceptual content of a sentence using the same neural system activated during actual perception

or action (Glenberg & Robertson, 2000; Pulvermüller, 2013). Just et al. (2004) found that low-

effort imagery activates the left temporal cortex and primary visual cortex, while high-effort

imagery requires generating a multidimensional internal structure of the event. Urrutia, Gennari &

De Vega (2012) identified brain areas sensitive to effort in action sentences, overlapping with

regions activated during physical effort tasks. These areas include the left inferior parietal lobe,

supramarginal and postcentral areas, associated with action planning. Prefrontal areas were

exclusively activated for counterfactual sentences. Action language involves events performed by

animate beings and entails physical effort (Trevisan et al., 2017). Research predominantly focuses

on physical content involving hand and arm movements (Moody & Gennari, 2010), but studies

have also examined mouth and leg movements (Tettamanti et al., 2006; De Vega, Urrutia & Riffo,

2007; Glenberg, 2007; Pulvermüller et al., 2005).

 7

The purpose of this article being to explore the degree to which embodiment is present during

processing of abstract language at varying levels of mental effort and in an indicative (real) and

counterfactual (unreal) linguistic context in older adults, and having described embodiment, as well

as mental effort, it’ll be necessary to further consider counterfactuality and representation theories

in the context of psycho- and neurolinguistics.

Counterfactual thinking involves contemplating alternative outcomes that did not occur but can be

imagined (Epstude & Roese, 2008). It is characterized by thoughts like "what if?" or "if only there

had been..." that explore different possibilities (Westby, 2016). Counterfactual reasoning entails

imagining alternatives to an event and shifting between real and counterfactual situations (Epstude

& Roese, 2008). The term "counterfactual reasoning" is used when individuals mentally bracket

what is known to be true and speculate about false possibilities (Urrutia & de Vega, 2012a).

Counterfactual language may interfere with action planning (De Vega & Urrutia, 2011) and

involves pre-motor regions that overlap with action execution networks (Boulenger, Hauk &

Pulvermüller, 2008). Counterfactuals elicit stronger activations and compete with each other,

requiring prefrontal inhibition or control processes (Beck, Riggs, & Gorniak, 2009; McNamara,

Durso, Brown & Lynch, 2003). Counterfactual reasoning involves prefrontal control processes that

decouple alternatives from actual states and engage competition between alternative interpretations

(Krueger, Barbey, & Grafman, 2009). Counterfactual comprehension activates medial prefrontal

regions and areas involved in resolving competition between alternative actions (Chen et al., 2009;

Goel, 2007; Simmonds, Pekar & Mostofsky, 2008; Speer et al., 2009; Yarkoni, Speer & Zacks,

2008). Urrutia et al. (2012) investigated whether the involvement of the body in counterfactual

sentences affected processing and found overlapping activations in premotor and parietal regions.

Counterfactual language has implications for cognitive processes like planning, causal reasoning,
 8

problem-solving, and decision-making (McNamara, Durso, Brown & Lynch, 2003). Older adults

with Parkinson's disease produce fewer counterfactual thoughts, suggesting a relationship between

counterfactual thinking and healthy cognitive aging (McNamara et al., 2003). Age-related

differences in memory processes during counterfactual reasoning have been observed, with older

adults showing vulnerability to false recall (Gerlach, Dornblaser & Schacter, 2014). Older adults

also exhibit lower inhibitory capacity for linguistic materials, impacting memory systems

(Radvansky, Zacks & Hasher, 2005). Counterfactual language imposes higher processing costs on

older adults, particularly working memory (Radvansky, Zacks & Hasher, 2005).

Multiple representation theories of abstract concepts have emerged as valuable and innovative

additions to current research (Gentner & Boroditsky, 2001; Barsalou, 2008). These theories suggest

that abstract concepts, although grounded in perception, action, and emotion, heavily rely on the

recreation of linguistic experiences (Paivio, 2007; Dove, 2009). According to proponents of these

theories, both sensorimotor and linguistically conveyed information contribute to the

representation of concrete and abstract concepts, but their distribution varies depending on the type

of concept (Binder & Desai, 2011). While perception and action information are more crucial for

concrete concepts, emotion and linguistically conveyed information play a greater role in abstract

ones (Kousta et al., 2011; Vigliocco et al., 2014).

The role of language in abstract concept representation is characterized differently by various

multiple representation approaches (Mahon & Caramazza, 2008; Dove, 2016). Some consider

language as a mere shortcut to access word meaning, while others assign it a more significant role

in enhancing thought processes or as a complex bodily and social experience (Pulvermüller, 2005;

Dove, 2016). This review focuses on four influential or potentially influential multiple

representation theories, including the Language and Sit Simulation (LSS) approach (Zwaan &
 9

Madden, 2005), Dove's representational pluralistic approach (Dove, 2009), and the Words as Social

Tools (PCHS) approach (Borghi & Cimatti, 2009).

The PCHS approach proposes that concepts are expressed through representations derived from

both perceptual/motor and linguistic experiences (Borghi & Cimatti, 2010). It emphasizes the

social dimension of word acquisition and considers both concrete and abstract words as social tools

for interaction (Borghi & Binkofski, 2014). The PCHS approach introduces four central principles,

which will be examined in light of recent evidence. These principles relate to the acquisition

modalities (Pulvermüller et al., 2014), conceptual representation in the brain (Barsalou, 2016),

embodied counterparts of different modalities (Glenberg, 2015), and the influence of social factors

on abstract word learning (Borghi et al., 2018). Research using fMRI (Binder et al., 2009), TMS

(Pulvermüller et al., 2005), and behavioral studies (Meteyard et al., 2012) has provided support for

the predictions made by the PCHS approach, highlighting the involvement of language processing

areas and motor systems in abstract concept representation.

Overall, these findings contribute to our understanding of the multimodal nature of abstract concept

representation and the role of language and social factors in this process (Kiefer & Pulvermüller,

2012; Barsalou, 2016).

 10

2. Material and methods

2.1. Participants

The sample consisted of 27 older adults from the city of Concepción, Chile (M = 70.34 years, SD

= 4.82, 15 women and 12 men). All of them were selected for showing a right hand preference,

according to the Edinburgh Handedness Inventory (Albayay et al., 2019). Only right-handed people

with a laterality coefficient greater than 60 were accepted. One participant was eliminated due to

the presence of excessive artifacts in her electroencephalography (EEG) recordings. In addition,

all participants passed the psychological tests applied. None of them evidenced alterations in the

central nervous system, abuse or problematic use of alcohol and other substances, learning or

memory problems, mental health problems or serious medical illness; and none were taking

medications known to have central effects. All participants had normal or corrected vision and were

native speakers of Chilean Spanish.

Before the experiment, each participant was evaluated with cognitive tests, screening and

psychopathological scales, in order to exclude participants with unwanted conditions (e.g.

depression, dementia, etc.). The Minimental State Examination (Folstein et al., 1983), with a

sensitivity of 93.6% and specificity of 46.1% (Quiroga et al., 2004), and the Yesavage Depression

Scale (Yesavage et al., 1982) were applied, with an internal consistency of 0.78 and construct

reliability of 0.87 (Angulo & Arias, 2011). In addition, a brief questionnaire was applied that

assessed functionality (especially in the sensorimotor plane). Participation was free and consented,

which was registered through a protocol validated by an Ethics Committee. The average duration

of the evaluation was 15 minutes.

 11

2.2. Materials and procedures

One hundred and sixty sentences corresponding to four types of semantics were elaborated: 40 high

mental effort factual sentences (HFS), 40 low mental effort factual sentences (LFS), 40 high mental

effort counterfactual sentences (HCS) and 40 low mental effort counterfactual sentences (LCS).

These sentences are exemplified in Table 1.

[suggested place for Table 1]

At the beginning of the task, the participants observed dashes and spaces on the computer screen.

The dashes corresponded to the letters of each word and the spaces to the segmentation places of

the sentence. Pressing a key unmasked the next segment and masked the one they just read. After

reading each sentence, participants were presented with a test word, which they had to decide

whether or not it was in the sentence they had just read by pressing the letter P (with a red sticker

on it) if their answer was affirmative or the letter Q (with a green sticker on it) if their answer was

negative.

In each experimental sentence, the intention was to record the event-related potentials (ERPs) of a

word (gerund/participle) and two syntagmas (direct object and adverbial). Technical problems

arose when measuring the direct object, so the results will refer to the gerund/participle as

POSITION 1 and the adverbial as POSITION 2. POSITION 1 reflects the effects associated with

the processing of the main verb of the sentence; therefore, it indicates the type of action performed

and implies, early on, the level of effort associated with it, as can be seen in the examples in Table

1. POSITION 2 captures the effects related to the closing of the sentence and the effects that are

incorporated and accumulated during its processing.

For example, in the sentence: “Now you are counting the trucks on the road”, a synchronous

signal was sent to the EEG corresponding to the beginning of the word “counting” (gerund) and

another signal corresponding to the syntagma “on the road” (adverbial). The purpose of recording
 12

both positions was to check the time course of the effort level processing in contexts of different

abstraction levels throughout the sentence. To analyze the electrophysiological signal at position 1,

200 ms prior to the analyzed position were designated as the baseline, and the total recording

window was extended to 1200 ms after the presentation of the stimulus. For position 2, 200 ms

were also designated as the baseline but the time window was extended to 1000 ms.

The experiment was self-administered using the E-Prime 2.0 Professional software.

2.3. Design

The experimental conditions were analyzed in a factorial design of two types of context (factual

vs. counterfactual) and two types of effort (low mental effort vs. high mental effort).

2.4. Electrophysiological record

A high-density EEG set-up was used, with 58 thin Ag/AgCl electrodes mounted on an elastic cap

that fitted the size of the head. Five electrodes were of the cup-type (10 mm in diameter) and were

placed in the ocular area: one under the left eye and another next to the ridge of the right eye to

measure eye movements. Two electrodes were placed in the mastoid area (under each ear), being

used as an average reference for the rest of the electrodes (monopolar recording) and one on the

forehead, which was used as grounding.

Electrode locations on the scalp were: FP1, FP2, F3, F4, C3, C4, P3, P4, O1, O2, F7, F8, T7, T8,

P7, P8, FZ, CZ, PZ, F1, F2, P1, P2, AF3, AF4, P5, P6, FC5, FC6, C5, C6, TP7, TP8, PO5, PO6,

FPZ, FCZ, CPZ, POZ, OZ, PO3, PO4, CP1, CP2, CP3, CP4, C1, C2, F5, F6, FC3, FC4, FC1, FC2,

CP5, CP6, PO7 and PO8. These locations followed the standard 10/20 electrode location system.

The inter-electrode impedance was kept below 5 KΩ. The biosignals were processed by a Neuronic

amplifier in a band between 0.02-100 Hz. In the ERP analysis, all those windows that contained

eye movements (EOG greater than 80ų) and those artifacts that contaminated the measurements
 13

such as noise of body movement or facial muscles, among others, were eliminated from the

individual averages. This procedure was done thanks to the filter that the Neuronic analysis

program performs automatically, but then each window was manually corrected to verify that the

elimination criteria were always the same and thus perform a stricter cleaning of each window. The

stimuli for the EEG recording system were presented through the E-Prime 2.0 Professional

software. The EEG signals were prepared for ERP analysis to reveal whether endogenous

components, such as the N400, distinguished between factual and counterfactual meaning, and

between high and low levels of mental effort. The choice of time windows for the ERP/EEG

analysis was based on theoretical and statistical criteria. Specifically, the nonparametric statistical

method of permutations (included in the Neuronic EP Workstation package) was applied to

estimate pairwise t-test comparisons between the levels of a variable at each data point (Khosla et

al., 2020). The ERP segments that reached the statistical criteria in the permutation tests were

selected as time windows for their analysis, according to the topographic distribution of the ERP

components described in the literature.

3. Results

200-600 ms window. There were significant effects on the set of 58 electrodes, with a significant

interaction Context x Effort F(1, 25) = 4.476, MSe = 4384465334.7, p < 0.05; η2 = 0.152; η2p/ƒ =

0.423; 1-β = 1.0 (Greenhouse-Geisser correction). In this time window, it is expected to find the

N400 component, with a negativity between 200 and 600 ms, greater in central-parietal sites, with

a bias towards the right hemisphere for written language (Issa et al., 2018). Thirty-four channels

were chosen as region of interest (ROI): C3, C4, P3, P4, P7, P8, P1, P2, P5, P6, C5, C6, PO5, PO6,

PO3, PO4, CP1, CP2, CP3, CP4, C1, C2, CP5, CP6, PO7, PO8, Fz, Cz, Pz, FPz, FCz, CPz, POz

and Oz, where significant interaction was found between the factors Context x Effort F (1, 25) =
 14

4.207, MSe = 330401238.1, p < 0.05; η2 = 0.144; η2p/ƒ = 0.41; 1-β = 1.0 (Greenhouse-Geisser

correction). This effect is shown in Figure 2.

[suggested place for Figure 2]

The medial channels were analyzed separately and a significant interaction was found between the

factors Context x Effort F (1, 25) = 4.238, MSe = 63211215.87, p < 0.05; η2 = 0.145; η2p/ƒ = 0.41;

1-β = 1.0 (Greenhouse-Geisser correction). In this time window, N400 was expressed mainly in the

posterior parietal and central areas and to a lesser extent in the parieto-occipital areas.

Figure 3 shows the voltage differences (amplitude) between the different experimental conditions.

It can be seen that the positivity of the high mental effort counterfactual predominates, followed

by the low mental effort factual. The topographic distribution of this temporal window is mainly

observed in the right central parietal areas, according to the topography of a classic N400. However,

the brain wave has a morphology with several voltage changes within the temporal range, which is

congruent with a typical wave of the study population.

[suggested place for Figure 3]

350-500 ms window. There were no significant effects on the set of 58 electrodes, finding only a

marginal trend (p = 0.08 > 0.05) in the Context x Effort interaction. As in the previous case, it was

expected to find a negativity around 400 ms (N400). Even though there are few and inconsistent

data on the effect of age on the N400 component, the literature suggests that healthy older adults

show a decrease in the effects of N400 compared to young adults (Joyal et al., 2020). Studies

indicate that the reduction in the N400 effect is due to smaller amplitudes of N400 for semantically

unrelated or incongruous stimuli (Joyal et al., 2020). There is also evidence that indicates that in

the older population the N400 component becomes more frontal when context information is

restricted to interpret or reinterpret a sentence (Wlotko & Federmeier, 2012), especially for those

older adults with greater verbal fluency (Wlotko et al., 2010), which is relevant given that at the
 15

sentence level to which this study was circumscribed, there was no contextual information before

the critical phrase.

Based on these empirical arguments, the ROI was extended to frontal areas, choosing the following

32 channels: F3, F4, C3, C4, P3, P4, T7, T8, F1, F2, P1, P2, C5, C6, TP7, TP8, PO5, PO6, PO3,

PO4, CP1, CP2, CP3, CP4, C1, C2, FC3, FC4, FC1, FC2, CP6, CP5, Cz, FCz and CPz. In these, a

significant interaction was found between the factors Context x Effort F(1, 25) = 4.220, MSe =

390132201.0, p < 0.05; η2 = 0.144; η2p/ƒ = 0.41; 1-β = 1.0 (Greenhouse-Geisser correction). Figure

4 shows the distribution of the effect on the semantics studied.

[suggested place for Figure 4]

The medial channels Cz, FCz and CPz were analyzed separately and a significant interaction was

found in them between the factors Context x Effort F(1, 25) = 4.797, MSe = 31794213.87, p <

0.05; η2 = 0.161; η2p/ƒ = 0.44; 1-β = 1.0 (Greenhouse-Geisser correction).

On the other hand, differences in amplitude were observed between the four experimental

conditions (Figure 5). Low-effort counterfactuals and high-effort factuals triggered the N400 effect.

In this temporal range this effect had a more frontal topographic distribution than in the previous

window.

[suggested place for Figure 5]

400-600 ms window. These results, in a way, were the extension of the results in the 200-600 ms

window. There were no significant effects on the set of 58 electrodes, finding only a marginal trend

(p = 0.058 > 0.05) in the Context x Effort interaction. The same channels were taken from the 200-

600 ms window (for the same reasons) and a significant interaction was found between the factors

Context x Effort F(1, 25) = 4.079, MSe = 333835235.8; p < 0.05; η2 = 0.14; η2p/ƒ = 0.42; 1-β = 1.0

(Greenhouse-Geisser correction). Figure 6 shows the direction of this effect and Figure 7 shows

the voltage amplitude differences.

 16

[suggested place for Figure 6]

Based on statistical criteria, the medial channels taken in this window were Cz, Pz, FCz, CPz, and

POz (not the eight medial channels in the 200-600 ms window). These were analyzed separately

and a significant interaction was found between the factors Context x Effort F(1, 25) = 4.396, MSe=

39918962.82, p < 0.05; η2 = 0.15; η2p/ƒ = 0.42; 1-β = 1.0 (Greenhouse-Geisser correction).

This negativity pointed in a direction similar to that of the findings of Zhu et al. (2018). These

authors found that N400 is delayed and has a smaller amplitude in older adults. It is important to

remember that the N400 component is not only modulated by the degree of semantic incongruity.

Among the various factors is the N400 concretion effect (Kounios & Holcomb, 1994). This effect

is typically seen in relation to concrete and abstract noun processing, with concrete nouns eliciting

improved frontal N400 responses compared to abstract nouns; and tends to be more pronounced

on the frontal scalp than on the parietal scalp (as opposed to the more central-parietal distribution

of the N400 semantic context effect for written words). The effect is also higher when word

processing goes beyond surface-level characteristics and when contextual constraints are weak

(Kounios & Holcomb, 1994).

The region of interest for this particular window is presented below, showing a right central-parietal

distribution. Due to the sinuous morphology of the brain wave, that we assume was characteristic

of the studied population, effects were found in different topographies in the N400 component.
 17

4. Discussion

This study explored the ERP components that are produced while processing sentences with

different levels of mental effort, in factual and counterfactual linguistic contexts, in elderly

population. The conceptual discussion of these results was carried out under the paradigm of

embodied and situated cognition (Schilhab, 2017; Winter, 2019). It is important to consider that

the analysis criteria will be framed in several aspects such as the characteristics of the population,

in terms of its cognitive decline, form of social and interpersonal interaction, as well as the

subjective emotional well-being of the elderly through the development of wisdom. This study also

provides elements for the analysis of both cognitive and affective processes involved in the

processing of these linguistic structures. The motor aspect will be used as a criterion and we will

delve in its interaction with the other analysis criteria described.

The results of each time window will be discussed below, following a sequence that allows a deeper

understanding of the components or effects found.

3.

4.

4.1. N400 component

Given the elicitation of the ERP N400 component in the three time windows, it is necessary to

discuss it in an integrated and sequential way. Based on the results found, the most striking thing

about this component lies in its latency. Negativity was expressed early at 200 ms and it was

extended to 600 ms.

In relation to these experimental antecedents, our results are novel as they show an onset 180 ms

earlier and 160 ms extended, compared to the typical time window in young people. The extension
 18

of the latency of the N400 has been interpreted as a consequence of the increase in the size of the

semantic networks related to the greater lexical-conceptual volume (Joyal et al., 2020). This

linguistic aspect is formed from the sensorimotor information present in the degree of mental effort,

producing greater negativity in actions of low factual mental effort than sentences of high mental

effort. On the other hand, the counterfactual also modulates the degree of mental effort, with low

mental effort being more negative than high mental effort.

Another notable aspect is that the sentences used in our study did not include words semantically

incongruous with the immediate linguistic context in the sentence final position. Our results show

that the N400 component was modulated, early and primarily, by the level of mental effort and the

counterfactual context, that is, embodied, emotional and linguistic aspects integrated with each

other in early processing zones (gerund).

If we observe the three moments of the N400 component in succession from its earliest stages of

onset (200 ms) to its latest (400 ms), we can observe that the high mental effort factual sentences

become more negative as the timeline advances. This implies that older adults process the

embodiment of the sentence as a less plausible event, semantically incoherent or pragmatically

little adjusted to the body-context relationship. On the other hand, in the high mental effort

counterfactual sentences, the positivity of the potentials decreased, probably due to the fact that the

reinterpretation of the events loses degrees of veracity due to the low semantic integration of the

events. There, older adults began to process the double counterfactual meaning, leaning towards

the reality of events. The N400 component, therefore, is modulated by linguistic and embodiment

factors.

A cognitive and embodied explanation for these effects can be found in the idea that high-effort

factual sentences are interpreted as events that are difficult to carry out, in relation to the effort that

it implies for an older person to generate mental actions that demand high mental energy levels.
 19

These types of actions are mentally represented or simulated as more demanding, and therefore

require more time and effort to be performed appropriately.

In parallel, the low effort factuals seem to offer greater opportunities to be simulated, represented

and semantically integrated, which is reflected in the positive voltages obtained. This facilitates

their processing and corroborates the idea that they function as more concrete elements than the

rest of the semantics expressed as experimental conditions of the study. On the other hand, it is also

striking that the high-effort counterfactual sentences were constantly more positive than the rest of

the sentences, reflecting that older adults feel, perhaps, stimulated to imagine challenging past

scenarios, which is probably experienced with a sensation of challenge, possibly accompanied by

positive emotions. The positivity when processing these sentences would therefore be linked to the

emotional salience that these sentences contain or trigger.

In relation to the brain regions that participate directly in the expression of the found N400

components, the early N400 effect was modulated by the right central and parietal zones. The

“standard” N400 component is regulated by mainly frontal and central zones. In contrast, late N400

appeared to be activated by brain activity in frontal, frontal-central, central, central-parietal, and

parieto-occipital areas.

The greater negativity of the high mental effort factual would imply that, by progressively building

the situation model of the sentence and reaching the critical zone where the gerund connects with

the direct object, the high effort in indicative grammatical mode tries to recruit personal

dispositions that the person unconsciously rejects or does not prefer, which would be interpreted

as implausibility or semantic/pragmatic incoherence. The opposite case occurs with the high mental

effort counterfactual, which continues to be salient, probably due to emotional and social factors.

The social aspect would be explained because the mandate contained in the sentence that begins

with “You should have...” is ultimately a social mandate oriented to the “should be” that comes
 20

from moral standards introjected from the culture and that could, eventually, imply the review of

behavior from the gaze of other people who judge or evaluate behavior.

Finally, it seems relevant to comment on these results as a whole, in relation to other studies. If the

assumption is accepted that low mental effort factual sentences are processed in a more concrete

way than high mental effort factual sentences, given their similarity to the everyday life of older

people (component N400) and that, on the other hand, the processing of high mental effort

counterfactuals is more salient for reasons related to their emotional load and double semantic

representation, the results in these time windows (component N400) are similar to those obtained

by Scorolli et al. (2011). These authors found that the processing of concrete or abstract words does

not only depend on the level of imaginability. It would also depend on how they are used in

linguistic-social contexts and different levels of self-relationship (Buccino et al., 2019; Scorolli et

al., 2011). This seems to be the case in our study since, despite the fact that mental verbs are

abstract, they can be embodied based on their high frequency of personal and social use, which

would provide evidence for the Theory of Words as Social Tools (Borghi et al., 2019).
 21

5. Conclusion

The present research provides evidence on the time course in which the aging brain processes

language. Both linguistic contexts -factual and counterfactual-, as well as the embodied parameter

of mental effort modulate the processing and participate with greater or lesser preponderance in

different time windows.

Counterfactual language seems to induce emotions that facilitate the processing of high effort,

whereas the low effort - counterfactual interaction represents a negativity that is associated with

the N400 component. The same goes for high-effort factual language. The appearance of an early

N400 is noteworthy, and could be related to both the high lexical knowledge of the elderly and the

failure of inhibitory processes, phenomena that seem to be stimulated by the high effort factual

semantics and low effort counterfactual semantics.

Older adults present a proportionally higher N400 for the high mental effort factual language than

for the low mental effort factual language, depending on the fact that they process the embodied

aspects of the sentence as less plausible, incoherent situations or less adjusted to their life context,

which coincides with the evidence that shows that language processing is facilitated when an action

is perceived in its usual context (Beauprez & Bidet-Ildei, 2019; Beauprez et al., 2020). On the other

hand, the counterfactual sentences behave in a different way, since their initial positivity

progressively declines given the reinterpretation of the double meaning of the events and leaning

towards the factual or real version of them. This effect could be of a linguistic-semantic nature,

probably modulated by emotions, which also constitute an embodied aspect.

Our analysis suggests that age-related declines in cognitive abilities, particularly those associated

with the mirror neuron system extending over the inferior frontal lobe, inferior parietal, and

superior temporal areas (Buccino et al., 2001), have the potential to impact various processes such
 22

as action observation and prediction, theory of mind, empathic resonance, motor resonance, and

sensorimotor simulation in the elderly. These findings have significant clinical relevance, as they

highlight the importance of considering the influence of age-related changes in the mirror neuron

system on language comprehension, specifically in relation to action language that incorporates

references (explicit or implicit) to actual effort, whether it be physical or mental.

Understanding how age-related cognitive changes affect language comprehension can have

important implications for the development of interventions and therapies aimed at improving

communication and linguistic abilities in older adults. By recognizing the specific linguistic

challenges that arise due to compromised mirror neuron system functioning, clinicians and

language therapists can tailor their approaches to better address the needs of older individuals.

Moreover, this research opens new avenues for investigating the potential benefits of targeted

cognitive training programs and rehabilitation strategies that aim to enhance the mirror neuron

system and its associated processes. By exploring ways to mitigate the effects of aging on action

language comprehension, we may be able to improve overall communication skills and quality of

life for older adults.

In summary, delving into the interplay between neural mechanisms, language processing, and

cognitive aging not only expands our theoretical understanding but also provides valuable insights

for clinical interventions and language therapies targeting older populations. Such endeavors hold

great promise in addressing the communication challenges faced by the elderly and improving their

overall cognitive well-being.

Declaration of interest: None.

 23

Funding: This work was supported by the Conicyt National Scholarship No. 21120957 of the

Doctorate in Linguistics of the University of Concepción, and the Project for the Acquisition and

Replacement of Scientific Equipment for Research 2022 "Cognitive Neuroscience Laboratory,

University of Bio Bio", code 2250315 AD/EQ of the University of Bío Bío, Chillán Campus. Chile.

Acknowledgements: We appreciate the funding granted by ANID/PIA/Basal Funds for Centers of

Excellence FB0003.

6. References

Albayay, J., Villarroel-Gruner, P., Bascour-Sandoval, C., Parma, V., & Gálvez-García, G. (2019).

Psychometric properties of the Spanish version of the Edinburgh Handedness Inventory in a sample

of Chilean undergraduates. Brain and Cognition, 137, 103618.

https://doi.org/10.1016/j.bandc.2019.103618

Allport, D. A., & Funnell, E. (1981). Components of the mental lexicon. Philosophical

Transactions of the Royal Society of London. B, Biological Sciences, 295(1077), 397-410.

https://doi.org/10.1098/rstb.1981.0148

Angulo, C. B. G., & Arias, A. C. (2011). Escala de Yesavage para Depresión Geriátrica (GDS-15

y GDS-5): estudio de la consistencia interna y estructura factorial. Universitas Psychologica, 10(3),

735-743. https://doi.org/10.11144/javeriana.upsy10-3.eydg

Araneda-Hinrichs, N. (2020). (New) realist social cognition. Frontiers in Human Neuroscience,

14, 337. https://doi.org/10.3389/fnhum.2020.00337

Beauprez, S. A., & Bidet-Ildei, C. (2018). The kinematics, not the orientation, of an action

influences language processing. Journal of Experimental Psychology: Human Perception and

Performance, 44(11), 1712-1726. https://doi.org/10.1037/xhp0000568

 24

Beauprez, S. A., & Bidet-Ildei, C. (2019). Perceiving a Biological Human Movement Facilitates

Action Verb Processing. Current Psychology, 38(5), 1355-1359. https://doi.org/10.1007/s12144-

017-9694-5

Beauprez, S. A., Blandin, Y., Almecija, Y., & Bidet-Ildei, C. (2020). Physical and observational

practices of unusual actions prime action verb processing. Brain and Cognition, 138, 103630.

https://doi.org/10.1016/j.bandc.2019.103630

Bidet-Ildei, C., Beauprez, S. A., & Badets, A. (2020a). A review of literature on the link between

action observation and action language: advancing a shared semantic theory. New Ideas in

Psychology, 58, 100777. https://doi.org/10.1016/j.newideapsych.2019.100777

Bidet-Ildei, C., Beauprez, S. A., & Boucard, G. (2020b). The link between language and action in

aging. Archives of Gerontology and Geriatrics, 90, 104099.

https://doi.org/10.1016/j.archger.2020.104099

Borghi, A. M., Barca, L., Binkofski, F., Castelfranchi, C., Pezzulo, G., & Tummolini, L. (2019).

Words as social tools: Language, sociality, and inner grounding in abstract concepts. Physics of

Life Reviews, 29, 120-153. https://doi.org/10.1016/j.plrev.2018.12.001

Borghi, A. M., Binkofski, F., Castelfranchi, C., Cimatti, F., Scorolli, C., & Tummolini, L. (2017).

The challenge of abstract concepts. Psychological Bulletin, 143(3), 263-292.

https://doi.org/10.1037/bul0000089

Buccino, G., Colagè, I., Silipo, F., & D’Ambrosio, P. (2019). The concreteness of abstract language:

an ancient issue and a new perspective. Brain Structure and Function, 224(4), 1385-1401.

https://doi.org/10.1007/s00429-019-01851-7

Cappuccio, M. L. (Ed.). (2019). Handbook of embodied cognition and sport psychology. MIT press.

Cowart, M. (2004). Embodied cognition. Internet Encyclopedia of Philosophy.

 25

Cox, A. (2016). Music and embodied cognition: Listening, moving, feeling, and thinking. Indiana

University Press.

de Vega, M. (2021). Revisitando la corporeidad del lenguaje narrativo. Revista signos, 54(107),

985-1003. https://doi.org/10.4067/S0718-09342021000300985

Folstein, M. F., Robins, L. N., & Helzer, J. E. (1983). The mini-mental state examination. Archives

of General Psychiatry, 40(7), 812. http://doi.org/10.1001/archpsyc.1983.01790060110016

Freud S. (1891). On aphasia: a critical study. International Universities Press.

Ionescu, T., & Vasc, D. (2014). Embodied cognition: challenges for psychology and education.

Procedia-Social and Behavioral Sciences, 128, 275-280.

Issa, M. F., Juhasz, Z. & Kozmann, G. (2018). EEG analysis methods in neurolinguistics: A short

review. Évfolyam, 2, 48-54.

Joyal, M., Groleau, C., Bouchard, C., Wilson, M. A., & Fecteau, S. (2020). Semantic Processing in

Healthy Aging and Alzheimer’s Disease: A Systematic Review of the N400 Differences. Brain

Sciences, 10(11), 770. https://doi.org/10.3390/brainsci10110770

Khosla, A., Khandnor, P., & Chand, T. (2020). A comparative analysis of signal processing and

classification methods for different applications based on EEG signals. Biocybernetics and

Biomedical Engineering, 40(2), 649-690. https://doi.org/10.1016/j.bbe.2020.02.002

Kounios, J., & Holcomb, P. J. (1994). Concreteness effects in semantic processing: ERP evidence

supporting dual-coding theory. Journal of Experimental Psychology: Learning, Memory, and

Cognition, 20(4), 804-823. https://doi.org/10.1037/0278-7393.20.4.804

Locke, J. (1847). An essay concerning human understanding. Kay & Troutman.

Loeffler, J., Raab, M., & Cañal-Bruland, R. (2016). A Lifespan Perspective on Embodied

Cognition. Frontiers in Psychology, 7, 1-6. https://doi.org/10.3389/fpsyg.2016.00845

 26

Marinescu, I., Lawlor, P., & Kording, K. (2018). Quasi-experimental causality in neuroscience and

behavioural research. Nature Human Behaviour.

Meteyard, L., Cuadrado, S. R., Bahrami, B., & Vigliocco, G. (2012). Coming of age: A review of

embodiment and the neuroscience of semantics. Cortex, 48(7), 788-804.

Miles, L., Nind, L., & Macrae, C. (2010). Moving through time. Psychol. Sci. 21, 222–223.

Miłkowski, M. (2013). Explaining the Computational Mind. MIT Press.

Muffato, V., Meneghetti, C., Di Ruocco, V., & De Beni, R. (2017). When young and older adults

learn a map: The influence of individual visuo-spatial factors. Learning and Individual.

Differences, 53, 114-121. https://doi.org/10.1016/j.lindif.2016.12.002

Portegijs, E., Karavirta, L., Saajanaho, M., Rantalainen, T., & Rantanen, T. (2019). Assessing

physical performance and physical activity in large population-based aging studies: home-based

assessments or visits to the research center? BMC Public Health, 19(1), 1-16.

https://doi.org/10.1186/s12889-019-7869-8

Quiroga, P., Albala, C., & Klaasen, G. (2004). Validación de un test de tamizaje para el diagnóstico

de demencia asociada a edad, en Chile. Revista médica de Chile, 132(4), 467-478.

http://doi.org/10.4067/S0034-98872004000400009

Royle, P., Steinhauer, K., Dessureault, É., Herbay, A. C., & Brambati, S. M. (2019). Aging and

language: Maintenance of morphological representations in older adults. Frontiers in

Communication, 4, 16. https://doi.org/10.3389/fcomm.2019.00016

Salas-Herrera, J. L. (2015). Lenguaje y envejecimiento desde una perspectiva corpórea. Paideia,

(57), 43-64. https://revistasacademicas.udec.cl/index.php/paideia/article/view/1495

 27

Salas-Herrera, J. L., Martínez, M. U., Araneda, R. M., & De Vos, M. V. (2020). Comprensión de

oraciones de esfuerzo en jóvenes y adultos mayores desde una perspectiva corpórea. Universitas

Psychologica, 19, 1-14. https://doi.org/10.11144/Javeriana.upsy19.coej

Samu, D., Campbell, K. L., Tsvetanov, K. A., Shafto, M. A., & Tyler, L. K. (2017). Preserved

cognitive functions with age are determined by domain-dependent shifts in network responsivity.

Nature Communications, 8(14743), 1-14. https://doi.org/10.1038/ncomms14743

Schilhab, T. (2017). Derived embodiment in abstract language. Springer.

Scorolli, C., Binkofski, F., Buccino, G., Nicoletti, R., Riggio, L., & Borghi, A. M. (2011). Abstract

and concrete sentences, embodiment, and languages. Frontiers in Psychology, 2, 227.

https://doi.org/10.3389/fpsyg.2011.00227

Shapiro, L. (2011). Embodied Cognition. Routledge.

Tirado, C., Khatin-Zadeh, O., Gastelum, M., Leigh-Jones, N., & Marmolejo-Ramos, F. (2018). The

strength of weak embodiment. International Journal of Psychological Research, 11(2), 77-85.

https://doi.org/10.21500/20112084.3420

Urrutia, M. & de Vega, M. (2012). Lenguaje y acción: Una revisión actual a las teorías corpóreas.

RLA. Revista de lingüística teórica y aplicada (50: 39–67).

Vukovic, N., Feurra, M., Shpektor, A., Myachykov, A., & Shtyrov, Y. (2017). Primary motor cortex

functionally contributes to language comprehension: An online rTMS study. Neuropsychologia,

96, 222-229. https://doi.org/10.1016/j.neuropsychologia.2017.01.025

Wernicke C. (1874). Der aphasische Symptomenkomplex. Cohn & Weigert.

Wilson, M. (2002). Six views of embodied cognition. Psychonomic Bulletin & Review (9, 4): 625-

636.
 28

Winter, B. (2019). Sensory linguistics: Language, perception and metaphor (Vol. 20). John

Benjamins Publishing Company.

Wlotko, E.W., Federmeier, K.D. 2012. So that's what you meant! Event-related potentials reveal

multiple aspects of context use during construction of message-level meaning. Neuroimage, 62(1),

356-366. https://doi.org/10.1016/j.neuroimage.2012.04.054

Wlotko, E. W., Lee, C. L., & Federmeier, K. D. (2010). Language of the Aging Brain: Event‐

Related Potential Studies of Comprehension in Older Adults. Language and Linguistics Compass,

4(8), 623-638. https://doi.org/10.1111/j.1749-818X.2010.00224.x

Yang, Y., Dickey, M. W., Fiez, J., Murphy, B., Mitchell, T., Collinger, J., Tyler-Kabara, E.,

Boninger, M., & Wang, W. (2017). Sensorimotor experience and verb-category mapping in human

sensory, motor and parietal neurons. Cortex, 92, 304-319.

https://doi.org/10.1016/j.cortex.2017.04.021

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, O., Huang, V., Adey, M., & Leirer, V. O. (1982).

Development and validation of a geriatric depression screening scale: A preliminary report. Journal

of Psychiatric Research, 17(1), 37-49. https://doi.org/10.1016/0022-3956(82)90033-4

Zhu, Z., Hou, X., & Yang, Y. (2018). Reduced Syntactic Processing Efficiency in Older Adults

During Sentence Comprehension. Frontiers in Psychology, 9, 243.

https://doi.org/10.3389/fpsyg.2018.00243
 29

7. Figure Legend

Fig. 1. Time sequence of an assay

Fig. 2. Average ROI 200-600 ms.

Fig. 3. Experimental conditions at C3 and C4 electrodes between 200-600 s.

Fig. 4. Average ROI 350-500 ms.

Fig. 5. Experimental conditions at C3 and C4 electrodes between 350-500 ms.

Fig. 6. Average ROI 400-600 ms.

Fig. 7. Experimental conditions at the Cz electrode between 400-600 ms.