Using the example of dispute on the status of linguistic competencies, being key for language researchers, the text embarks on the issue of alternative interpretations of so-called 'empirical data'. Groundlessness of a rhetoric of 'scholarliness' and 'objectivity', as can be found in Steven Pinker's texts, is indicated. Pinker has several times formulated '‘empirical arguments' testifying to innateness of a language instinct. He has referred to data from various areas, including research on cognition in babies or language acquisition. The present article aims, among others, to show that there are alternative ways of interpreting such data, which indeed can be reconciled with diverse concepts of language. An exemplary opposing concept is Michael Tomasello's approach. Reference is also made to an article by Marshall M. Haith, which explicitly points out to a risk of overinterpretation of laboratory test results concerning cognition in babies.