Abstract
SummaryScientific realism is here made equivalent to the referentiality of scientific language. A clear distinction of meaning and reference is advocated and certain ‘symptoms' of referentiality in scientific language are stressed. It is then shown that contemporary scholars stressing the contextual determination of meaning, the meaning variance and theory‐ladenness of all terms in scientific theories, often fail to recognize that an independent ‘stable’ core of the meaning still exists. This allows for theory comparison and is witness that science investigates reality, provided one is aware that different kinds of reality are investigated by different sciences. Naive realism amounts to disregarding that no access to reality is given outside meaning and language, naive antirealism amounts to overlooking that language refers to something different from language itself