Both　text　and　figure　are　two　basic　visual　forms　to　express　the　meaning　of　information.　However,　the　relationship　between　the　semantic　processing　of　text　and　of　figure　is　still　not　clear　among　cognitive　scientists.　This　study　investigated　whether　semantic　processes　in　the　two　forms　of　expressing　the　same　meaning　of　information　have　shared　and　separate　neural　systems　using　eye-tracking　and　functional　neuroimaging.　In　the　eye-tracking　study,　the　results　suggest　that　the　text　and　figure　have　no　significant　difference　on　the　reading　effect,　which　supports　the　common-code　theory.　In　the　fMRI　study,　using　conjunction　analysis,　there　were　greater　common　regions　as　a　shared　semantic　network　involving　the　frontal　lobe,　parietal　lobe,　temporal　lobe　and　limbic　system,　which　suggests　that　the　greater　shared　activiations　as　a　common　semantic　system　for　the　content　of　both　text　and　figure;　the　direct　comparisons　of　the　two　forms　showed　that　the　text　was　more　significant　activations　of　striate　and　prestriate　cortex,　whereas　figure　was　more　significant　activations　of　the　parahippocampal　gyrus　and　extrastriate,　which　indicate　that　form-specific　activations　in　visual　cortex　are　involved　in　the　initially　semantic　processing.　Taken　together,　our　results　provide　the　converging　evidence　for　that　the　text　and　figure　mainly　share　a　common　semantic　system,　and　also　have　the　form-specific　semantic　system　in　initially　processing.　©　2011　IEEE.