Table 1.
Overview of functional magnetic resonance imaging (fMRI) studies investigating involvement of inferior frontal, sensorimotor and inferior parietal systems in syllable perception.
| No. | Study | Stimuli features investigated | Phonetic | Task | Button presses | Sparse imaging | Analyses | Baseline | Activation/Decoding found in … | ||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Prefrontal areas | Motor areas | Somatosensory and inferior parietal areas | |||||||||
| 1 | Benson et al. (2001) | 15 C/VC/CVC syllables | n/a | none | never | yes | univariate | non-speech tones | left BA 9, 10 | left BA 6 | left SMG (BA 40) |
| 2 | Wilson et al. (2004) | /pa/, /gi/ | n/a | none | never | no | univariate | rest/silence | not reported | ventral (v) BA 4, 6 | right SMG (BA 40) |
| 3 | Wilson and Iacoboni (2006) | 50 consonants embedded between two /α/ vowels | n/a | none | never | no | univariate | rest/silence | not reported | v BA 4, 6 | not reported |
| 4 | Szenkovits et al. (2012) | monosyllabic pseudowords | n/a | one-back repetition detection task | 10% of trials | no | univariate | non-speech buzzes | not reported | not reportedd | not reported |
| 5 | Grabski et al. (2013) | 9 vowels | n/a | none | never | yes | univariate | rest/silence | BA 44, left BA 45 | right BA 6 | not reported |
| 6 | Pulvermüller et al. (2006) | /pæ/, /tæ/, /pI/, /tI/ | Place | none | never | yes | univariate, ROI-baseda | matched noise stimuli | not reported | left v BA 4, 6 (differential activation of lip vs. tongue regions) | not reported |
| 7 | Raizada and Poldrack (2007) | /ba/, /da/ | Place | detect occasional quieter stimulus | 6% of trials | yes | repetition adaptation | n/a | left middle frontal cortex (amplification of response to stimulus pairs differing in place) | not reported | left SMG (amplification of response to stimulus pairs differing in place) |
| 8 | Myers et al. (2009) | /da/, /ta/ | Voicing | detect occasional high-pitched stimulus | 37.5% of trials | yes | repetition adaptation | n/a | left inferior frontal sulcus (release from adaptation only for stimuli differing in voicing) | not reported | not reported |
| 9 | Lee et al. (2012) | 10 CV syllables on /ba/-/da/ continuum | Place | detect occasional quieter stimulus | 11% of trials | yes | MVPA (searchlight) | n/a | left BA 44 (decoding of place) | left pre-SMA (decoding of place) | not reported |
| 10 | Chevillet et al. (2013)b | /da/–/ga/ continuum | Place | dichotic listening (detect in which ear the sound persisted longer) | always | yes | repetition adaptation | n/a | not reported | left BA 6 (release from adaptation for stimulus pairs differing in place of articulation) | not reported |
| 11 | Du et al. (2014) | /ba/, /ma/, /da/, /ta/ | Place | active syllable identification (4-AFC) | always | no (but scanner noise attenuation by 25 dB) | MVPA (searchlight) | n/a | Insula/Broca’s area (decoding of place)—at low/moderate noise levels onlye | left v BA 6 (decoding of place)—at low noise levels onlyf | left inferior parietal lobule (decoding of place)—at low noise levels only |
| 12 | Arsenault and Buchsbaum (2015) | 16 CV syllables | Place, manner, voicing | gender identification task | always | no | MVPA (ROI-based) | n/a | not reported | not reported | left subcentral gyrus (decoding of place) |
| 13 | Evans and Davis (2015) | /ba/, /da/, /ma/, /na/, /ab/, /ad/ | Place, manner, phoneme identity, CV structure (CV vs. VC) | one-back repetition detection task | 8% of trials | yes | MVPA-based RSA (searchlight)c | rest/silence | not reported | left precentral gyrus (decoding of syllable and phoneme identity and CV structure) | left postcentral gyrus (decoding of syllable identity) |
| 14 | Correia et al. (2015) | 24 CV syllables | Place, manner, voicing | none | never | yes | MVPA (searchlight) | n/a | IFG (decoding of place/manner) | right inferior precentral gyrus (decoding of place) | postcentral gyrus, SMG (decoding of place/manner) |
| 15 | Arsenault and Buchsbaum (2016) | 8 CV syllables: /ba/, /pa/, /va/, /fa/, /da/, /ta/, /za/, /sa/ | Place (manner/voicing not analyzed) | detect occasional blank trials | 11% of trials | no | univariate and MVPA (both ROI-based) | n/a | not reported | not reported | left postcentral gyrus (decoding of place) |
aDifferential activation of subregions (lip vs. tongue) in left BA 4, 6 depending on place of articulation (lip vs. tongue; see Figure 2 Top). bSee also Alho et al. (2016) for similar results from MEG. cRepresentational similarity analysis (RSA) used to disentangle acoustic and phonological similarity. dNot reported in whole-brain analysis; but note that in a ROI-based analysis a correlation between working memory abilities and motor/premotor activation in speech perception was found. fIn dorsal BA 6 (PMC) decoding was still observed at higher noise levels (see Figure 3), but this might reflect classification of responses rather than speech perception (see main text for discussion). eDue to the active identification task on every trial, this might reflect decision-related processes or response selection/preparation (see main text for discussion).