<p>Biological species are widely regarded as fundamental units for generalizing about the natural world. Despite ongoing debates over their conceptual nature, they remain central to research and practice across the life sciences. Species are treated as products of evolution, existing independently of human perception, and discovered through investigation. Their existence is formalized in descriptions and scientific names, with publication in recognized outlets serving as the foundation of identification. The question of how many species exist holds major scientific and political significance. In the context of climate change and biodiversity loss, it is more than academic: species counts are vital indicators for ecosystem assessment and form the basis of conservation policy. Yet this question is not purely empirical. Species numbers are epistemic constructs shaped by historical, methodological, and normative factors. While formal naming under taxonomic codes is treated as discovery, the number of named species is uncertain, and the number of undescribed species remains largely unknown—an urgent issue amid accelerating extinctions. New molecular and AI-based taxonomic methods aim to speed discovery but also challenge traditional notions of knowing a species. Examining species counting—from early catalogues to modern data infrastructures—reveals as much about scientific and cultural frameworks as about biodiversity itself. This paper is a contribution to the Special Issue "Artificial Researchers and ScientificDiscoveries".</p>