Kohler’s observation that most people match pseudoword “maluma” to curvy objects and “takete” to spiky objects represented the well-known example of sound symbolism—the idea that link between sound and meaning of words was not entirely arbitrary. This study was aimed to examine the existence of sound symbolism in natural language and to consider the potential role of some aspects of experimental design and stimuli features which had not been considered in experimental studies so far. Three experiments were done in order to explore the influence of visual information on language processing. Visual lexical decision task with the sharp-sounding and soft-sounding verbal stimuli presented within the spiky and curvy frames was used. Reaction time analysis in these three experiments highlighted additional aspects of visual and language processing which influence the potential interplay of these two processes. As results revealed, when visual information preceded presentation of verbal material for approximately 1000 ms or when visual and verbal material were presented simultaneously, the processing was being delayed and the interactions of these two processes occurred. The pattern of obtained results gave further support to the idea of sound symbolism as pre-semantic phenomenon and the hypothesis that the effect emerged from very early stages of language processing.