Establishing propositional truth-value in counterfactual and real-world contexts during sentence comprehension: differential sensitivity of the left and right inferior frontal gyri.

TL;DR: The fMRI study investigated the neural circuits that are sensitive to the propositional truth-value of sentences about counterfactual worlds, aiming to reveal differential hemispheric sensitivity of the inferior prefrontal gyri tocounterfactual truth- Value and real-world truth- value.

About: This article is published in NeuroImage.The article was published on 2012-02-15 and is currently open access. It has received 32 citations till now. The article focuses on the topics: Counterfactual conditional & Counterfactual thinking.

Summary

Introduction

• Language is a computationally remarkable, uniquely human system, not to mention their principal and most efficient means of communication.
• These findings testify to the fact that an extended network of brain regions is sensitive to sentence truth-value, even if participants are not explicitly instructed to establish truth-value.

Participants

• Twenty-eight right-handed students (14 males, mean age=22.9 - years) participated in this study for monetary compensation.
• All participants were native Spanish speakers, had normal or corrected-to- normal vision, and gave written informed consent.
• Participants had no neurological or psychiatric disorders, nor had they seen the materials before in an earlier experiment.
• Four participants were excluded from the final analysis due to excessive movement during the experiment.

Development and pretest of materials

• Materials were identical to those used by Nieuwland and Martin (2012), and were selected from 133 Spanish sentence quadruplets with two counterfactual and two real-world sentences (see Table 1).
• Counterfactual sentences described hypothetical consequences of common-knowledge historical events not having taken place, whereas real-world sentences described actual consequences of these events.
• The authors first established the expectedness of critical words.
• Twenty students of the University of the Basque Country completed one of two lists with one version of each item truncated before the critical word.
• Twentyfour different students evaluated one of four counterbalanced sentence lists containing only one condition per quadruplet, and decided whether the sentences were true (1=False, 7=True), skipping any they could not evaluate.

Experimental procedure

• While inside the scanner, participants read sentences presented via back-projection onto the middle of the screen, and would view the stimuli via a mirror attached to the head coil.
• Moreover, while Tesink et al. reported effects of truth-value of single sentences, the Menenti et al. study involved multi-sentence stories rather than single sentences, and the reported coordinates in that study might thus not be optimal for the comparisons in the current study.
• FMRI data acquisition, preprocessing and analysis Imaging took place on a 3-T MR scanner (Siemens TrioTim) with echoplanar imaging capability.
• Each subject then viewed one of the four counterbalanced sentence lists with the sentence trials and fixation trials, across six functional runs.
• The explanatory variables were modeled as a fixed response (box-car) waveform temporally convolved with the canonical HRF along with its temporal derivative (Friston et al., 1998), while controlling for serial correlations.

Region-of-interest analysis

• Given the a priori hypothesis about the role of the IFG, a region-ofinterest (ROI) analysis was performed using the Marsbar toolbox (Brett et al., 2002) by extracting average parameter estimates per condition and per subject for 2 LIFG ROIs and their right hemisphere counterparts.
• These ROIs were based on the results of a related study that reported activations within different subregions of the left and right IFG (pars orbitalis/triangularis; BA 45/47) for world knowledge violations in healthy adults during sentence processing (Tesink et al., 2011; see also Tesink et al., 2009, for similar results).
• A 2(Context: counterfactual, real-world) by 2(Propositional truth-value: true, false) by 2(hemisphere: left, right) repeated measures analysis of variance was performed on the extracted parameter estimates per ROI.

Whole-brain analysis

• The whole-brain analysis did not reveal any significant clusters for the interaction between Context and Propositional truth-value.
• Consistent with the absence of a significant Context by Propositional truth-value interaction effect, the false minus true contrast elicited similar activation increases for counterfactual and real-world sentences (see Table 3 and Fig. 2).
• These clusters were located in left and right inferior frontal gyrus (LIFG and RIFG, BA 45 and BA 47) extending into the middle frontal gyrus (MFG, BA 8/9), left middle temporal gyrus (MTG, BA 21), medial parts of the superior frontal gyrus (SFG, BA 6/8), and the left inferior parietal lobule (IPL, BA 40).
• Consistent with the main effect of context, the contrast (Counterfactual False>Real-world False) elicited a significant cluster in LMTG (k=440, peak voxel coordinates [−48 −32 4]), as did the contrast (Counterfactual True>Real-world True) (k=1421, peak voxel coordinates [−50 −24 −14]).

ROI analysis

• Differences in parameter estimates corresponding to the effect of truth-value (false minus true sentences) per sentence type and ROI are plotted in Fig. 1b.
• This differential effect of truth-value in the right BA 47 ROI for the two sentence types was of a quantitative nature rather than a qualitative: the effect of truth-value was statistically significant for each sentence type, but it was stronger in counterfactual sentences (F(1,23)=6.83, pb .001) than in real-world sentences (F(1,23)=2.21, p=.037).
• This difference was driven mainly by a large response to counterfactual false sentences, which was larger than the response to real-world false sentences (F(1,23)=3.59, p=.002), while the responses to counterfactual true and real-world true sentences did not differ (F(1,23)= 1.06, p=.31).

Discussion

• This fMRI study investigated the neural circuits that are sensitive to the propositional truth-value of sentences about counterfactual worlds or about the real world, and aimed to reveal hemispheric differences between the inferior prefrontal gyri in processing counterfactual truth-value and real-world truth-value.
• No such pattern was found for the left and right BA 45 ROIs.
• The current larger sensitivity of the RIFG to counterfactual truth-value than to real-world truth-value may thus reflect increased semantic processing due to the semantic complexity of the counterfactual false sentences.
• The larger effect in the RIFG for counterfactual truth-value is consistent with right hemisphere activations associated with comprehending contextual and figurative meaning and with complex sentences and discourse level processing (e.g., Kuperberg et al., 2006; Stringaris et al., 2006; Xu et al., 2005 see Jung-Beeman, 2005; Mason and Just, 2006).
• First of all, these differences occurred before the critical words, and their effects were accounted for by separate.

MFG

• Contexts may have been the use of historical counterfactual conditionals.
• Whereas Menenti et al. created novel and unfamiliar fictional stories, alternative endings to known historical events may be more easily computed, for example because relevant information is also part of their real-world knowledge (e.g., the fact that the Soviets were also making substantial progress in landing somebody on the moon at the time that the USA managed to do so).
• In the current study, counterfactual and real-world sentences elicited similar effects of truth-value in the left middle temporal gyrus.
• These results may reflect increased semantic retrieval of relevant information in long-term memory (e.g., of the correct information) from medial temporal lobe, governed by the inferior prefrontal gyri (e.g., Bookheimer, 2002; Hagoort et al., 2009; Lau et al., 2008).
• Interestingly, activity increases in this region have been reported for syntactic errors but not for semantic anomalies (they may even result in relative deactivations; e.g., Hagoort et al., 2004; Nieuwland et al., 2007, 2011; Tesink et al., 2009).

Acknowledgments

• This research was supported by a Plan Nacional research grant from the Spanish Ministry of Science and Innovation (grant number PSI2010-18087) to MSN.
• Many thanks to Javi Miqueleiz for help with stimulus construction, to Natalia Barrios and Larraitz Lopez for help with data collection, and to two anonymous reviewers and Andrea Eyleen Martin for comments on an earlier version of this manuscript.

Figures

Table 1 Example sentences and approximate translations with average truth-value rating and cloze value of the critical word for each condition. This table is adapted from Nieuwland and Martin (2011).

Table 2 Schematic representation of the regressors corresponding to the different sentence parts.

Table 3 Results from the Whole-brain analysis.

Fig. 1. (left graph) ROIs in the current study. Two 10-mm sphere centered at MNI coordinates [−44 29 12] (BA 45, top) and [−36 26–8] (BA 47, bottom) and their right hemisphere equivalents. (right graph) Effect of propositional truth-value in counterfactual sentences (dark gray bars, ‘CF’) and real-world sentences (light gray bars, ‘RW’) in four ROIs (left/ right, BA 45/47). Displayed are the false-minus-true difference score (and 95% confidence intervals) in average beta parameter value per sentence type and per ROI. (*pb .05).

Fig. 2. Effect of propositional truth-value in counterfactual sentences (A) and in real-world sentences (B). Pair-wise comparisons for false minus true sentences across all subjects (thresholded at P≤0.001 uncorrected). Labels are provided only for significant clusters (corrected for multiple comparisons with a cluster-level FDR P≤0.05). LIFG = Left inferior frontal gyrus. RIFG = Right inferior frontal gyrus. SFG = Superior frontal gyrus. MFG = Middle frontal gyrus. MTG = Middle temporal gyrus. IPL = Inferior parietal lobule.

Frequently asked questions

Q1. What are the contributions in "Establishing propositional truth-value in counterfactual and real-world contexts during sentence comprehension: differential sensitivity of the left and right inferior frontal gyri" ?

This paper investigated the role of the left inferior frontal gyrus for counterfactual reasoning about what is in fact false as if it were true.