Transitional Fossils

"There are no transitional fossils" is among the most widely repeated claims in creationist literature. It is also testable, which is the only reason it is worth a page. Either museum collections contain specimens with intermediate anatomy or they do not.

What follows are specific fossils, the journals that published them, what makes each one intermediate, and — for the best cases — the prediction that came before the discovery.

One clarification first, because the term invites a persistent misunderstanding. A transitional fossil is not a half-formed creature struggling toward a goal. It is a functioning animal that happens to combine features now split between two later groups. Tiktaalik was not trying to become a tetrapod. It was a fish that was good at propping itself up in shallow water.

The Prediction That Came First#

The strongest evidence in science is a prediction made before the data exists, because it cannot be retrofitted.

In the late 1990s Neil Shubin, Edward Daeschler, and Farish Jenkins wanted a fossil showing the fish-to-tetrapod transition. Rather than searching randomly, they reasoned from the theory: such an animal should appear in shallow freshwater deposits from the Late Devonian, roughly 375 million years ago, since fully aquatic lobe-finned fish were known from earlier and clear tetrapods from later. They found exposed rock of the right age and depositional environment on Ellesmere Island in the Canadian Arctic and went digging there.

It took four field seasons. In 2004 they found Tiktaalik roseae, published in Nature in 2006. It has gills and scales like a fish. It also has a flattened skull, eyes on top of its head, a neck — fish do not have necks — and, inside its fin, a shoulder, an elbow, and a wrist.

The theory named the age of the rock, the type of rock, and the anatomy. All three could have been wrong. A prediction that specific, confirmed by a targeted search, is difficult to account for as coincidence or as post-hoc interpretation.

Cases With Good Intermediate Series#

Whales#

This is the clearest sequence we have, partly because whales are large and their bones fossilize well, and partly because the transition is recent enough that the rocks are abundant.

Pakicetus (~50 Ma) was a wolf-sized land animal with an ear region — specifically the involucrum, a dense bony wall of the middle ear — found in no group except whales. Ambulocetus (~49 Ma) was amphibious, with large hind feet and a body plan suited to crocodile-style ambush. Rodhocetus (~47 Ma) had shortened limbs and a tail adapted for swimming. Basilosaurus and Dorudon (~40–34 Ma) were fully aquatic and retained tiny, functionless hind limbs complete with knees and toes.

Molecular evidence independently placed whales inside the even-toed ungulates, closest to hippos — a conclusion from DNA that the fossil ankle bones later confirmed. Two lines of evidence with no methodological overlap, agreeing.

Birds and dinosaurs#

Archaeopteryx, described in 1861, has feathers and a wishbone alongside teeth, a long bony tail, and clawed fingers. It was for a century the sole example, which made it easy to dismiss as a one-off oddity.

Since the mid-1990s, the Yixian Formation in Liaoning, China has produced hundreds of feathered dinosaurs — Sinosauropteryx, Caudipteryx, Microraptor, Anchiornis — spanning simple filaments through downy tufts to fully vaned, asymmetric flight feathers. Some are preserved well enough that melanosome shape and distribution have been used to reconstruct plumage color.

Human ancestry#

Sahelanthropus (~7 Ma), Ardipithecus ramidus (~4.4 Ma), Australopithecus afarensis (~3.2 Ma, including the Lucy skeleton), Homo habilis, Homo erectus, Homo heidelbergensis, Homo neanderthalensis. Brain volume roughly triples across this series while the pelvis and foot reorganize for upright walking, and the two changes do not happen at the same time — bipedality comes first, by a wide margin.

The Laetoli footprints in Tanzania, dated to 3.66 million years ago, record a fully bipedal gait pressed into volcanic ash long before brains enlarged. That ordering was not obvious and it is not what most early researchers expected.

Tetrapod to amniote, and mammal-like reptiles#

The synapsid series — Dimetrodon, therapsids, cynodonts, early mammals — documents the jaw joint reorganizing and two bones of the reptilian jaw migrating into the mammalian middle ear. Some intermediate forms have a double jaw joint, with both the old and new articulations present simultaneously in the same skull. This is a direct answer to the claim that such a transition could not occur gradually: the intermediate stage is not hypothetical, it is a fossil using both configurations at once.

Common objections#

"Every new fossil creates two new gaps"

True, and it follows from the structure of the claim: any two forms have a space between them, so filling one produces two smaller ones.

Fossilization is rare. It requires rapid burial in suitable sediment chemistry, and the rock must then survive hundreds of millions of years without being subducted, metamorphosed, or eroded. A sparse record is what would be expected.

What the theory predicts is not completeness but consistency: whatever is found must fall in the right order. In roughly two centuries of stratigraphy on every continent, no one has found a mammal in Devonian rock, a flowering plant in the Cambrian, or a modern bird beneath the dinosaurs that preceded it. Every catalogued specimen was an opportunity for the order to break.

"Some transitional fossils turned out to be frauds"

Two cases are usually cited. Archaeoraptor was published by National Geographic in 1999 and was a composite assembled by a fossil dealer. Piltdown Man was a forgery that stood from 1912 until 1953.

Both are real, and National Geographic's handling of Archaeoraptor was a genuine editorial failure — it was published in a popular magazine before peer review.

The detection is the relevant part. Archaeoraptor was exposed within months by paleontologists using CT scanning, before formal scientific publication. Piltdown was suspect for much of its life precisely because it fit the emerging pattern poorly, and was exposed by fluorine dating.

Neither is part of the current evidence base, and neither was caught by critics of the field.

"Tiktaalik was overturned by older tetrapod trackways"

Accurate. Trackways from Zachełmie in Poland, dated to roughly 395 million years ago, predate Tiktaalik by about 18 million years. Answers in Genesis and the Discovery Institute have both cited this as a refutation.

It does revise the timeline. Tiktaalik is a late-surviving member of a lineage rather than the first of its kind.

It does not affect the anatomy or the prediction. Shubin's team searched Devonian freshwater deposits of a specific age because that is where an intermediate should occur, and found an animal with a mobile neck, ribs capable of supporting weight, and fin bones corresponding to the tetrapod limb. The prediction concerned what would be found in rock of that age, and the finding stands.

"Punctuated equilibrium admits the fossils are missing"

Eldredge and Gould's 1972 proposal concerned the rate of change: speciation tends to occur rapidly in small isolated populations and then stabilize, which makes transitions geologically brief and less often sampled.

That is an explanation of why intermediates are less common than a strictly gradual model predicts, not a concession that none exist.

Both authors objected repeatedly to creationist use of their work. Gould described the statement that there are no transitional fossils as a completely false characterization of the record, and wrote that creationists had quoted him selectively.

"These are just distinct created kinds that happen to look similar"

Answers in Genesis classifies whales, for instance, as a created kind unrelated to land mammals.

The difficulty is the ordering. The intermediate forms are not scattered through the column; they appear in the correct stratigraphic sequence, with Pakicetus below Ambulocetus below Rodhocetus below Basilosaurus, over roughly ten million years.

Separately created kinds have no reason to appear in an order matching increasing similarity to modern whales. The synapsid series makes the same point: some intermediates carry a double jaw joint, using the reptilian and mammalian articulations simultaneously in one skull.

What the evidence shows#

The fossil record contains extended intermediate series for several major transitions, including fish to tetrapod, land mammal to whale, dinosaur to bird, and reptile to mammal.

In each case the intermediates occur in the predicted stratigraphic order. Tiktaalik was found by searching rock of a specific age chosen because an intermediate should be there.

The record is incomplete and will remain so, because fossilization is rare. What it is not is inconsistent: no specimen has been found outside the interval the theory allows, across two centuries of collection.

Genetic evidence bearing on the same conclusion is in Chromosome 2 and Endogenous Retroviruses.