Mutation, selection, gene duplication, regulatory change, and population divergence can be studied directly. Major transitions in the distant history of life cannot be repeated in their original historical setting.
Researchers therefore reconstruct ancient transitions by combining several kinds of evidence: comparative genomics, developmental biology, phylogenetic relationships, fossils, molecular similarities, experimental studies, and the distribution of traits among living organisms.
This means that evidence for a major historical transition is normally cumulative. No single experiment is expected to reproduce the entire transition.
Research into evolutionary novelty distinguishes the appearance of novel attributes from their later refinement, diversification, and ecological success. Developmental and genomic studies can sometimes identify components and regulatory systems associated with major anatomical differences, while fossils and phylogenies provide historical context.
The further back a transition occurred, the less complete its surviving record may be. Historical reconstruction can identify plausible sequences and mechanisms without recovering every mutation or every intermediate population that once existed.
For which major transitions is the connection between known mechanisms and historical evidence particularly detailed? Where do important gaps remain between demonstrated mechanisms and reconstructed events?
Intelligent Design should distinguish between an event that has not been directly observed and an event for which no evidence exists. Historical sciences routinely infer past events from surviving evidence. The relevant design question is whether the available evidence and known mechanisms provide an adequate explanation for the particular transition being investigated.