Ever since Wegener, many writers have advanced the idea of Earth expansion. Alfred Wegener's continental drift theory provoked intense debate during the 1920s; it also spawned a flurry of books and articles, written mainly in Russian and German, advocating Earth expansion. The expansion concept "lies implicit within Wegener," but in the 1920s, most geologists had trouble enough with wandering continents; few were prepared to deal with an expanding earth. Besides, everyone 'knew' that the Earth is contracting. Earth expansion seemed so farfetched that those men who did take it seriously assumed that no one else had even considered the idea. Unaware of one another's work, each was convinced of his own priority. According to H.W. Menard, Earth expansion was 'discovered' no fewer than eight times between 1920 and 1960.

During the golden age of marine geology in the 1950s, when continental drift became respectable again after decades in contempt, interest in the expansion hypothesis was also revived. Earth expansion beneath the oceans during the last 200 million years could account for many of the new and surprising facts about the ocean floor – its relative youthfulness, the shallow sediments, and the worldwide rift system. For a time, a few North American geologists, including Bruce Heezen and J. Tuzo Wilson, flirted with expansion. But by the late sixties, most of these erstwhile supporters had abandoned expansion in favor of the new theory of Plate Tectonics

The expansion hypothesis remains a minority view in geophysics but it still has many supporters, especially among Earth scientists living in the southern hemisphere. Its most vigorous defender is S. Warren Carey, Emeritus Professor of Geology from the University of Tasmania and a past president of the Australian and New Zealand Association for the Advancement of Science. Though little known to the public outside Australia, Carey is widely regarded as a pioneer in modern Earth science. He came to believe in continental drift during his college days in the 1930s, long before it became fashionable. Unfortunately, anti-drift sentiment in those days was so strong that Carey felt compelled, for the sake of his D.Sc., to delete a controversial section on continental drift from his dissertation. But Carey's youthful convictions were vindicated by professional experience. Working as an oil geologist in New Guinea, "where the drifting land masses become most self-evident", and serving there during World War II, Carey became a confirmed 'mobilist' at a time when 'fixist' dogma was still firmly entrenched.

In the early 1950s, Carey developed a convection theory of continental drift that was strikingly similar to the Hess-Dietz model of seafloor spreading proposed a decade later. But Carey's convection theory was never published: the referees at the American Geophysical Union rejected his paper as "naive and unpublishable." In the mid-fifties, Carey verified, with great accuracy and without the benefit of a computer, the close 'fit' between South America and Africa – a full decade before Bullard. Many of Carey's 'naive' ideas from the forties and fifties became the orthodoxy of the sixties and seventies. "Any of my older students will confirm that what I taught them in the mid-1950's differed little from what the Kuhnian revolution [in Earth science] had brought. They found little new in the 'new' global tectonics." Though Carey may not have received the credit he is due, his work has not been completely ignored. Tuzo Wilson recalls that "the turning point" in his conversion to continental drift "was the receipt of a copy from Sam Carey of his mimeographed paper from the [1956] Tasmanian [drift] symposium." And Bullard never denied Carey's priority.

Carey embraced the expansion hypothesis in the mid-fifties after attempting, without success, to reconstruct Pangaea, the ancestral supercontinent, on a globe of modern dimensions. No matter how the continents were arranged, huge 'gaping gores' would appear between regions with known geological connections. "I could make satisfactory sketches like Wegener's classic assembly [Carey recalls], but never accurately on the globe, or a rigorous projection. Starting from the assembly of Africa and South America, [...] a yawning gulf* appeared between Indonesia and Australia [which] belonged together." Years of frustration tempted Carey to abandon his ambitious project. "A crucial link seemed to be missing from the global synthesis." But in the end, Carey's zeal for accuracy paid off, and the solution was surprisingly simple. "It was not my method that was at fault, but my implicit assumption that the Earth of Pangaea was the same size as the Earth today. The assembly of Pangaea was not possible on a globe of present radius, but on a smaller globe, ...these difficulties vanished." Carey had found the missing link.

The gaping gores reveal huge gaps in current thinking about the earth. Though psychologically comforting, the assumption of a constant-sized Earth is fraught with paradox – contradictions of geological fact – when attempts are made to reassemble Pangaea. In addition to the gaping gores, (subsequently verified by Owen in his Atlas of Continental Drift) Carey also discovered many other geological impossibilities when Pangaea is assembled on a globe of present size. These include the Pacific Paradox, the Arctic Paradox, the India-Gondwana Paradox and the Double Equator Paradox, to name but a few. All of these enigmas disappear when the continents are assembled on a smaller globe. And they helped convince Carey that the Earth had expanded.

None of the positive evidence marshaled in support of Plate Tectonics is incompatible with Earth expansion. The expansion hypothesis can easily account for continental displacement, the creation of new oceanic crust, transform faults, the concentration of earthquakes at plate boundaries, and many other well known geological phenomena. But which explanation is correct? One way of judging two competing theories is by measuring their predictive power; and in this respect the expansion hypothesis demonstrates its superiority. Crucial facts that Plate Tectonics must 'explain' – such as the youthful age of the Pacific floor, intraplate seismic activity (e.g. the New Madrid quake of 1823), or the Paleozoic geological links between India and Asia – are predictable consequences of Earth expansion. (Indeed, the relative displacement of the continents is perhaps its most compelling prediction.) Conversely, many important predictions deduced from Plate Tectonics – such as the existence of accretionary prisms of sediments scraped off subducted plates and the alleged compressional origin of oceanic trenches – have been refuted.

Plate subduction is believed to occur primarily in the Pacific at the deep trenches in mid-ocean and along the continental margins. According to the plate theory, seafloor spreading in the Atlantic, Indian and Arctic oceans causes the plates to converge on the Pacific where they collide and overlap. When this occurs, one of the plates is subducted beneath the other and it descends back into the earth's interior (where it is 'made mantle again') along the seismically active Benioff-Wadati Zone. Subduction into the trenches supposedly eliminates crust as quickly as seafloor spreading creates it. For every square mile of new crust created by seafloor spreading, an equal amount of old crust must be consumed by subduction. Without subduction, or some other crust disposal mechanism, the earth's surface area would increase and the Earth would expand.

When the 'conveyor-belt' model of seafloor spreading cum subduction was first conceived in the 1960s, it was confidently predicted that the trenches would exhibit signs of compression, caused by the convergence of lithospheric plates and the subduction of thousands of square miles of oceanic crust. But this prediction proved wrong. "The seafloor spreading hypothesis may, for some geologists, require compression in the vicinity of trenches, but the data require horizontal tension." According to H. W. Menard, a specialist in Pacific geology, "the most troublesome aspect of the sea-floor spreading hypothesis was the absence of direct evidence of convergence. There was no problem if the Earth was expanding, but if it was not, enormous areas of old oceanic crust had to be plunging into the mantle along the line of oceanic trenches. It was generally expected that the sediment in trenches would show signs of this violent phenomena, but none could be found." In trench after trench, the sediments turned out to be completely undisturbed; and there were no outcrops from the subducted plates. Menard and his fellow oceanographer Maurice Ewing were mystified by all this. "Neither of us believed for a moment in an expanding earth, so we were left with a puzzle." But in the end, both men endorsed the troublesome subduction hypothesis. Perhaps they believed they had no choice.

The continental drift debate lasted, off and on, for nearly half a century. The reality of drift was proven geologically in the 1930s by the South African geologist Alexander L. Du Toit, but it was not until the 1960s that the majority of geologists (and the Americans in particular) were finally converted. Given the prolonged and agonizing struggle over continental drift – an idea with overwhelming evidence in its favor – it is rather surprising that subduction – an untested if not untestable hypothesis with practically no empirical support – was accepted sight unseen, and practically overnight. Geology is a science renowned for its conservatism. Why, then, did so many geologists convert so quickly to an idea so radical and unfounded? The reason is simple: subduction was the lesser of two evils. The alternative – Earth expansion – was "anathema to most geophysicists," and it still is. Evidently, it is easier to believe in a super-efficient crust disposal mechanism that consumes oceanic crust at precisely the same rate as it is formed without leaving any evidence in its wake, than it is to consider the simpler (albeit still unexplained) alternative of Earth expansion. The universal rejection of Earth expansion led inexorably to the reification of plate subduction. Today, subduction is portrayed as though it were The Gospel Truth, unblemished by critical scrutiny. Unfortunately, subduction contains far more gospel than truth. Carey is more blunt. Subduction, he writes, "is a myth that exists only in the minds of its creators."

Another point of contention between Plate Tectonics and Earth expansion is orogenesis – the origin of mountains and fold belts. According to the plate theory, orogenesis is a compressional phenomenon: folding results from crustal foreshortening caused by the collision of two lithospheric plates. But Carey denies this; he attributes orogenesis to vertical uplift and gravitational collapse on an expanding sphere. Once again, Carey has the geological evidence on his side. The Himalaya, the very archetype of compressional plate tectonics, "could not have been born of collision nor of subduction, but resulted from vertical uplift." And in the Andes, "the fundamental tectonic style is extensional (east-west)." Unfortunately, when faced with the choice between fact and doctrine – the fact of extension and vertical uplift versus the doctrine of compression and crustal foreshortening – most Anglo-American plate theorists opt for the latter. Naturally, they have devised all sorts of ad hoc explanations for these anomalous facts, a tactic that appalls many field workers and which philosophers of science consider symptomatic of a 'degenerating research program.' Noting that most German and Russian-speaking geologists have abandoned this doctrine, Carey dismisses compressional orogenesis as an "English-language obsession" – a legacy of Newton's Earth contraction theory.

The pioneers of the 'new global tectonics' have helped revolutionize Earth science. Unfortunately, they are still encumbered with many obsolete dogmas. The adherence to dogma in the face of contrary evidence, and the proliferation of ad hoc hypotheses to account for that evidence, is not at all unusual in science – especially during a scientific revolution. Consider, for example, Galileo. A courageous advocate of the new Copernican system, the great Italian scientist could never bring himself to accept Kepler's evidence that the orbits of the planets are elliptical and not circular. Still entranced by the ancient Greek idea of heavenly perfection, Galileo believed that the planets move around the Sun in perfectly circular orbits. Unable or unwilling to abandon this archaic doctrine – despite the evidence against it – Galileo, like Ptolemy and Copernicus before him, was compelled to invoke epicycles 'to save the appearances' of planetary motion. A similar situation exists in geology today. Advocates of Plate Tectonics have contributed enormously to the 'Copernican Revolution' in Earth science, but because they are unable or unwilling to abandon the dogma of a constant-size earth, these latter-day Galileos must invoke subduction – the modern equivalent of the Galilean epicycle – 'to save the appearances' of continental motion on an Earth of constant size.

Most geologists assume implicitly that the earth's size has not changed. But if that assumption is wrong – and the geological evidence suggests that it is – then any theory based on that assumption must also be wrong, regardless of its popularity or the extent of its corroboration.