The neural correlates of reference to the past

Background Recently, Bastiaanse et al. (2011) identified past time-reference morphology as a particular source of the prominent verb inflection problems in agrammatism. Using a verb form with past time-reference requires additional processing, because a link has to be made to the event in discourse—this in contrast to non-past time-reference. Discourse-related electrophysiological processing differences were demonstrated in non-brain-damaged speakers earlier (Bos et al., 2013). As tense is not equivalent to time-reference, this theory applies to all discourse-linked verb forms, regardless of tense. For example, reference to the past can be made by simple past (‘V-ed’), or by an auxiliary in present tense plus a participle, the periphrastic past (‘has+V-ed’). For Dutch aphasic speakers both types of past time-reference are more difficult than simple present (Bos & Bastiaanse, 2014). Sahin et al. (2006) found left inferior frontal gyrus (IFG) and supplementary motor area (SMA) involvement in processing grammatical morphology for simple past versus simple present tense. We investigated if and where discourse-linked past time-reference is reflected in additional brain activation for periphrastic verbs forms that differ in time reference, but not in tense of the auxiliary. Methods We carried out an event related fMRI-study with 20 healthy Dutch speakers (mean age 24). They were presented with a picture denoting the target verb action. Subsequently, a written sentence had to be completed with non-past or past time-reference. One or two series of dots indicated single or periphrastic target verb forms, respectively. Contrasts were (1) simple past versus simple present tense (skated–skates) and (2) periphrastic past versus future time-reference (has skated–will skate). We analysed DARTEL-normalised functional scans of response preparation from picture offset to response onset in SPM8 with a factorial design (time-reference x simple/periphrastic). Results/discussion Analysis (see Figure 1) showed greater BOLD signal change for both types of past over non-past time-reference bilaterally frontal—the SMA and frontal superior medial gyrus. We ascribe this to increased difficulty in selecting past time-reference inflection, hence, discourse linking. We found significantly increased LIFG activation for simple past over simple present, but not for periphrastic past over future (see Figure 1C). LIFG might be taxed for past tense inflection, but not past time-reference specifically. Although periphrastic past and future differ in time-reference, both verb clusters have a present tense auxiliary. Our results align with a fundamental difference between past and non-past time-reference, and are in accordance with past time-reference processing difficulties of aphasic individuals.