Aim
We review the works of three pre‐Darwinian authors—Alexander von Humboldt, Augustin de Candolle and Alphonse de Candolle. Contrary to what other historiographers have claimed, we think that their works were not merely descriptive, nor did they adhere strictly to the inductivist model. Furthermore, the patterns described by these authors are still being studied extensively in biogeography.
Method
We analyse the main concepts provided by these authors on the spatial patterns of biodiversity, and discuss how they served to construct the conceptual framework of contemporary biogeography.
Results
In the early 19th century, secular naturalists began postulating explanations supported by empirical evidence and based on physical causes. The studied authors were extremely cautious when proposing hypotheses, considering that any statement not based on empirical evidence was mere conjecture. When we analyse the way they developed explanations, it seems evident that they were rather closer to deductive thought than to the orthodox inductivist model. With the rise of Darwinian evolutionism, naturalists lost interest in the spatial patterns of life and began constructing dynamic models where dispersal of individual lineages from northern centres of origin was proposed to explain their current distribution. Paradoxically, even though the evolutionary paradigm soon proved its great explanatory power, it simultaneously promoted a large number of studies in which speculation prevailed.
Main conclusions
Even though the practices and methods of current biogeography have changed substantially, the studied patterns are still the same: (1) geographical gradients of richness and biotic replacement, (2) spatial gradients of functional groups, (3) geographical variation of the phenotype, (4) geographical patterns of expansion‐differentiation of single lineages and (5) biogeographical homology. The studied authors not only described rigorously biogeographical patterns, but also proposed explanations about them and even sometimes tested them based on empirical evidence, going beyond the prevalent inductivist canon.