Flat Gaps

"There is special need for a term for this non-evident type of unconformity in which the contact is a simple bedding plane, since its recognition is commonly subjective." — Carl Dunbar and John Rodgers, Principles of Stratigraphy, 1957, the book that coined the word¹

"With deposition, there is no gap, because sedimentation continues.  With erosion, one would expect abundant channeling and the formation of deep gullies, canyons and valleys;  yet, the contacts are usually 'nearly planar.'" — Ariel Roth, Origins, 1988²

"Throughout the area of this study the Chattanooga shale rests on a peneplain that is notable for its extent and degree of perfection.  In the Tennessee area alone the shale rests on no less than 23 different sedimentary formations" — Louis Conant and Vernon Swanson, U.S. Geological Survey Professional Paper 357, 1961³

Summary

Stand at the rim of the Grand Canyon and look at the line between the Muav Limestone and the Redwall Limestone.  It is a bedding plane.  Nothing is scoured into it.  Nothing is weathered on it.  No soil sits between the two.  Geologists put the missing time at more than 100 million years.⁴  A young-earth geologist asks the obvious question:  where did the time go?  If the surface sat exposed, it should be gullied.  If sediment kept arriving, there is no gap.

That is the flat-gap argument.  Its technical statement is Ariel Roth's, in Origins in 1988.²  It is a good argument.  It also has an answer, and the answer is a set of measurements.

Eight conclusions, stated first.

1. The observation is real, and the discipline named it before any creationist used it. "Paraconformity" was coined by Carl Dunbar and John Rodgers in Principles of Stratigraphy in 1957, for the case "in which the beds are parallel and the contact is a simple bedding plane."¹ In the same paragraph they wrote that "its recognition is commonly subjective."¹ Edwin McKee of the U.S. Geological Survey described the Grand Canyon's pre-Mississippian contact as "a flat, even surface … in many places … obscure."⁵ Anyone who calls the flat gap a creationist invention is arguing with the textbook that named it.
2. The creationist argument runs one test. The standard list has eight. Dunbar and Rodgers set out the physical evidence for a break in 1957: surface relief including caves; weathering, an old soil profile, or a rusty or calichified zone; truncated joints, faults and dikes; pebbles of the lower unit in the base of the upper; lag gravels; and "concentrates of phosphate, manganese, iron, or glauconite nodules or grains."¹ Roth's papers test relief only. The words "hardground," "paleosol," "condensed," "glauconite," "phosphorite" and "bioturbation" appear nowhere in Roth 1988. Andrew Snelling's 1992 checklist for the Muav-Redwall contact has seven items, and every one is structural or textural.⁶ The other tests were never run.
3. A subcrop map settles it. A photograph cannot. The base of the Chattanooga Shale in Tennessee is one of the two type paraconformities named in Dunbar and Rodgers, and a 91-page Geological Survey monograph mapped it.¹ The shale "rests on no less than 23 different sedimentary formations … ranging in age from Middle Ordovician to Middle Devonian," and the angular discordance is "rarely as much as 1°."³ That surface truncates twenty-three formations and still looks like a bedding plane at every outcrop. So flatness at outcrop scale says nothing about how much rock is missing. A 1° bevel removes about 17 metres of section for every kilometre travelled. No eye sees 1° on a cliff face.
4. Roth's erosion arithmetic uses the wrong measurement. His 61 metres per million years is Judson and Ritter's 1964 figure for sediment yield out of United States drainage basins.² A bare low-relief rock surface needs the outcrop rate. That rate comes from a compilation of 1,599 cosmogenic beryllium-10 measurements: outcrops average "12 ± 1.3 m Myr−1," the median is "5.4 m Myr−1," and rates "do not exceed 140 m Myr−1."⁷ Roth's predicted 3,000 metres at the Grand Canyon's 100-million-year gap becomes about 540 metres. That does not make the problem vanish. 540 metres is still far more relief than the contact shows. The rest of the answer comes from the setting.
5. Limestone dissolution gives the hardest number in the debate. Ten published measurements of limestone surface lowering span 0.003 to 0.100 millimetres per year.⁸ At the fastest rate ever measured, one Flood year removes 0.1 millimetre of limestone. The valleys cut into the top of the Redwall average 75 metres deep, with solution caves inside the limestone. That needs 0.75 to 25 million years. Grant a thousandfold speed-up for hot acidic Flood water and one year still gives 10 centimetres.
6. Time can be measured on a perfectly flat submarine surface, by chemistry. Roth asks whether a submerged gap would simply be protected from erosion.⁹ It would. It would also grow minerals that keep time. An evolved glaucony grain takes more than 10^4-10rs to form on a starved seafloor, and the Fe(III)-smectite to glauconite step alone runs "~10^4-106 years."¹⁰ A surface with no erosion on it can still carry a record.
7. Hardgrounds are the sharpest test of order, and neither side has a clock. Mark Wilson of the College of Wooster argues that borings widen downward as the animal grows, cut across earlier borings already filled with cemented sediment, and are overgrown by encrusters that had to settle as larvae on rock that was already hard.¹¹ Tas Walker answers that hardgrounds are pre-Flood seafloor debris cemented during dewatering, and that the borings are compaction moulds that squeezed the animals out "like a stone from a squeezed plum."¹¹ Wilson's own concession belongs here too: "Forming a hardground probably takes years, but we do not yet have any direct way to estimate the duration of the process."¹¹
8. The leading young-earth stratigraphic model agrees with Roth, and it has put a number on the gap. Timothy Clarey and Davis Werner run the Institute for Creation Research's megasequence programme across roughly 3,000 stratigraphic columns. They define a megasequence as a package "bounded top and bottom by inter-regional unconformity surfaces" that "can be recognized by their bounding erosional surfaces."¹² That is a rule for finding a boundary in a column, not a claim that much rock was removed there. On erosion they stand with Roth and Michael Oard: "the consistent internal stratigraphy of each megasequence testifies against significant erosion."¹² At the Muav-Redwall contact they write that the two limestones "may have been deposited one upon the other in a matter of days or weeks with little erosion in between," and they give the flat top of the Muav as the evidence.¹³ So the young-earth camp is not split on these surfaces. The answer this article gives Roth applies to Clarey too, and Clarey's calendar makes his version the easier one to test.

The thing that decides it

The flat-gap argument treats a contact as a thing you look at.  Every measurement that settles one treats it as a thing you map, date, or dissolve.  None of those results is visible from a canyon rim.  None is answered by a photograph.

Part 1: What a paraconformity is, and who named it

This article is about contacts inside the Phanerozoic column.  The basement contact is a different surface with different evidence, and it has its own article.

An unconformity is a gap in a rock sequence.  Dunbar and Rodgers sorted the cases into four types and named the fourth.  They restricted "disconformity" to a parallel contact that is "an old erosion surface of appreciable relief," and introduced "paraconformity" for the case "in which the beds are parallel and the contact is a simple bedding plane."¹  Two things follow.  The creationist literature is right about both.

A paraconformity is defined by what you cannot see.  Dunbar and Rodgers said so:  "There is special need for a term for this non-evident type of unconformity in which the contact is a simple bedding plane, since its recognition is commonly subjective."¹  Roth's definition is the same one, shorter:  "if the contact is not visible or there is no evidence of erosion, it is called a paraconformity."²  Nobody is misrepresenting anybody here.

The hiatus can be enormous.  The type example in the 1957 book is Beargrass Quarry near Louisville, Kentucky.  There, Middle Silurian limestone meets Lower Middle Devonian limestone at a bedding plane.  "This hiatus is represented by more than 3,000 feet of strata in parts of the Appalachian trough."¹  Three thousand feet of section somewhere else.  At Beargrass Quarry, a line.

The creationist literature leaves out the rest of the same chapter.  Seven pages later Dunbar and Rodgers list the physical evidence a geologist looks for at a break.  The list is longer than "relief or nothing."  It runs:  "pronounced irregularities, mounds, crevices, or even caves";  "a fire clay, an old soil profile, or a rusty, silicified, or calichified zone";  "the abrupt truncation of structural features such as joints, faults, or dikes";  "pebbles from the beds below in the basal layers above";  lag gravels;  and "concentrates of phosphate, manganese, iron, or glauconite nodules or grains (though such concentrates may be evidence only of extremely slow deposition and not of a break in the record)."¹

That last parenthesis answers Roth's dilemma, thirty-one years before Roth wrote.  A surface can record time by depositing almost nothing, very slowly, for a very long while.  It does not have to erode.  It does not have to bury.  It can starve.

They also warned against reading the size of a hiatus off the look of a contact:  "It must not be inferred, however, that the angle of discordance indicates the relative importance of the hiatus."¹  That is the discipline's own position, in its founding text.  The rest of this article defends it with numbers.

Part 2: The argument at full strength

Roth's 1988 paper in Origins is the technical original.  Every later version traces to it.  He states it in three sentences.

"The difficulty with the extended time proposed for these gaps is that one cannot have deposition, nor can one have much erosion.  With deposition, there is no gap, because sedimentation continues.  With erosion, one would expect abundant channeling and the formation of deep gullies, canyons and valleys;  yet, the contacts are usually 'nearly planar.'"²

Roth's strongest authority is Norman Newell, who put the problem as plainly as any creationist has:  "A remarkable aspect of paraconformities in limestone sequences is general lack of evidence of leaching of the undersurface.  Residual soils and karst surfaces that might be expected to result from long subaerial exposure are lacking or unrecognized."¹⁴  Newell added:  "The origin of paraconformities is uncertain, and I certainly do not have a simple solution to this problem."¹⁴  Creationist writers have quoted that second sentence ever since.

The examples go beyond the Grand Canyon.  Roth lists a 200-million-year gap at Palo Duro Canyon in Texas;  a 10-million-year gap at Dead Horse Point, Utah, which he says covers about 250,000 square kilometres;  the top of the Bulli Coal in the Sydney Basin;  and the Morcles Nappe in the Rhone valley.²

His 2009 restatement in the Journal of Creation attaches a number to every photograph.  At roughly 30 metres of erosion per million years, the Grand Canyon's three gaps should show 180, 420 and 3,000 metres of lowering, and the Morrison-Dakota gap 1,200 metres.⁹  That is the version worth arguing with.  Roth has committed to a predicted depth of erosion at every named contact.

The sharpest form of his case is the Dakota Sandstone:

"The top of the Morrison Formation had to remain extremely flat for 40 Ma in order to accommodate the thin Dakota Formation, which is spread for many thousands of square kilometers immediately above it."⁹

The Dakota "averages only about 30 m in thickness, but it is spread over some 800,000 km2 in the western United States."⁹  Forty million years of nothing, then a thin blanket over most of a continent's interior.  A skeptic should feel the force of that.

Roth also makes the claim that turns the argument from local to global:  "One must keep in mind that evidence of time missing in one locality suggests the absence of that time over the entire globe."²  In the same paper he concedes the fact that most weakens him:  "Rarely, at these planar surfaces, there is some evidence of deeper erosion, sometimes even several hundred meters of erosion.  However, the general pattern shows little erosion."²

Two other authors carry the argument.  Andrew Snelling's 1992 article picks one locality on the North Kaibab Trail where the Devonian Temple Butte is absent and the Redwall sits directly on the Muav.  He lists seven features he would expect at a 155-million-year break, and reports that "not even one of the seven expected features listed above can be seen at this supposed boundary."⁶  Michael Oard's 2016 note in the Journal of Creation takes the same surface, adds Miall's missing-time work to it, and concludes that the evidence "further implies that the claimed millions of years are imaginary."¹⁵

Steven Austin's 1994 Grand Canyon: Monument to Catastrophe deserves separate credit.  It quotes its mainstream sources honestly, including the two sentences that damage its own case most.  Austin prints Dunbar and Rodgers' statement that fossils are "the final and the only criterion that gives quantitative results for the large unconformities," framing sentence and all.¹⁶  He prints Ronald Blakey's flat contradiction of the whole flat-gap argument:  "Unfortunately, presence or absence of conglomerate, relief, or contrast in lithology has little to do with the extent or amount of time missing at a given unconformity."¹⁶  Then he answers it rather than hiding it:  "the evaluation of the time significance represented by erosion between strata is tied ultimately to one's assumptions."¹⁶

That is the argument.  Its authors state it fairly.  It rests on a real observation.  It is wrong.

Part 3: The erosion rate is the wrong number

One sentence does the arithmetic:

"Present rates of erosion for the United States average about 6.1 cm/1000 yrs (Judson & Ritter 1964), which is 61 m/Ma.  In 10 Ma one would expect 600 m of erosion, etc."²

Judson and Ritter measured how much sediment American rivers carry out of their drainage basins.  That number folds in uplifted mountains, steep slopes, soft soil, farmland and gullies.  It does not measure how fast a bare, flat rock surface goes down.

Since 1988 that second quantity has been measured directly.  Cosmogenic beryllium-10 builds up in quartz within about two metres of the surface.  Its concentration therefore records how fast that surface is being stripped.  Eric Portenga and Paul Bierman ran 1,599 such measurements through one calculator.  The result splits exactly along Roth's error:

• Drainage basins: "218 ± 35 m Myr−1 (n = 1149)," median 54.⁷
• Bare outcrops: "12 ± 1.3 m Myr−1," median "5.4 m Myr−1," and "Compiled outcrop erosion rates are slow and do not exceed 140 m Myr−1."⁷

Roth's figure sits inside the basin distribution, which is the one that does not apply.  Swap in the outcrop median and his 3,000 metres at the Grand Canyon's longest gap becomes about 540 metres.  His claim that "Kangaroo Island should have been eroded down 6 km in 200 Ma" becomes roughly 1 kilometre at the global outcrop median.²

This does not rescue the mainstream position on its own.  Pretending it does would be a bluff.  540 metres of lowering is still far more relief than a Muav-Redwall contact shows.  Two further things have to be true, and they are.

First, the rate applies only while the surface is exposed.  A cratonic platform at or just below sea level, under a shallow sea, does not erode subaerially at any rate at all.  McKee's explanation of the Grand Canyon flatness sits in the same paragraph creationists quote for the flatness:  "nowhere … evidence of large uplift" that would dissect the region.⁵  No uplift, no dissection.

Second, the missing systems are missing only locally.  Two hundred and fifty kilometres west of the Grand Canyon, at the Nevada Test Site, the same interval holds roughly 3,400 feet of Ordovician and 1,415 feet of undifferentiated Silurian and Devonian dolomite.¹⁷  The Ordovician and Silurian seas existed and deposited a kilometre of rock.  They did it in the subsiding basin to the west and skipped the starved craton edge to the east.  A paraconformity of that kind records non-deposition.  Non-deposition on a flat, submerged platform leaves nothing to photograph.

That is Newell's own answer.  The creationist literature stops quoting him a paragraph before he gives it.  On the same p. 364 he names the hypothesis he favours:  "a concept of long protracted stillstand of relative sea level adjacent to very low land surfaces.  Yet this seems to be the most probable explanation of the prevalent absence of unequivocal evidence of leaching and oxidation at paraconformities."¹⁴  On his last page he writes the sentence the whole flat-gap argument has to get past:  "The fact that paraconformities are difficult or impossible to recognize by casual inspection is not proof that they do not exist."¹⁴

Part 4: The Chattanooga bevel, and why the test is a map

The cleanest refutation of "flat means no time" is a Geological Survey monograph on one of the two paraconformities Dunbar and Rodgers named as type examples:  the base of the Chattanooga Shale in central Tennessee.¹

Louis Conant and Vernon Swanson mapped it across the Nashville Dome in 1961.  Their p. 13 is the datum:

"Throughout the area of this study the Chattanooga shale rests on a peneplain that is notable for its extent and degree of perfection.  In the Tennessee area alone the shale rests on no less than 23 different sedimentary formations (pl. 3), as mapped by Wilson (1949, pl. 2), ranging in age from Middle Ordovician to Middle Devonian.  The older beds were tilted so slightly before Chattanooga time that in most outcrops the shale appears to parallel the older rocks;  at a few places an angular discordance is visible (pl. 5A) but is rarely as much as 1°."³

At any single outcrop the shale looks parallel to what it rests on.  Across the region it truncates twenty-three formations spanning about 85 million years of section.  Both statements are true of the same surface.

The reason is geometry.  A slope of 1° drops about 17 metres over one kilometre.  A bevel at that angle removes a kilometre of section in 60 kilometres of travel, while every roadcut still shows a contact that looks like a bedding plane.  One degree is the most Conant and Swanson report anywhere, and they see it only at a few places, so the real figure is smaller.  Flatness at outcrop scale says nothing about the size of a hiatus.  It says something about the local dip, and the local dip is under one degree.

Conant and Swanson describe what that took:  "Limestones, sandstones, and shales were beveled and planed during what must have been a time of unusual crustal stability until a wide area was reduced to an almost featureless plain barely above sea level."³

There is a second reason to use this monograph.  Under the shale, in places, lies a silty clay residuum that looks like weathered limestone.  It would be easy to call it a fossil soil and add it to the case.  Conant and Swanson do not.  They argue that most of it probably formed after the modern land surface was cut, and they give their reasons:  "nothing resembling the A and B soil horizons of pedologists has been identified";  "the residuum has not been found in the freshest deep highway cuts";  "the residuum is not present in any of the 75 drill cores that have penetrated the lower contact of the shale."³

Those are Tas Walker's criteria, applied by the Survey against its own convenience, thirty-two years before Walker published them.¹⁸  Concede it.  The sub-Chattanooga residuum is not a paleosol, and this article does not use it as one.  The bevel does not need it.

One young-earth author has written about this contact.  John Morris of the Institute for Creation Research called the shale "underlain by an amazingly flat, featureless erosional surface, with no significant high or low areas where deposition took place," and read that as "rapid sheet erosion."¹⁹  That describes the outcrop.  It does not touch the map.  Morris does not mention the twenty-three formations.  No other young-earth treatment found for this article mentions them either.  The word "Chattanooga" does not occur in Roth 1988, Roth 2009, Snelling 1992, Oard 2016, Reed and Oard 2018, or Austin 1994.  A Flood model that wants this surface has to plane 85 million years of section flat to within one degree across central Tennessee, then lay a black shale on it without disturbing the bevel, inside a year.

Part 5: Three clocks that run on a flat surface

The flat-gap argument assumes that a surface with no erosion on it has no record on it.  Three lines of evidence say otherwise.  None of them needs a channel, a gully or a soil.

Limestone dissolves at a measured rate

Limestone in contact with rainwater or soil water goes into solution.  How fast is a measured quantity.  Vicky Banks and Peter Jones measured it on a Derbyshire railway cutting driven in 1863, using the relief between resistant chert bands and the limestone around them:  "The mean rate of dissolution has been calculated from the time elapsed since the cutting was driven (1863) as 0.035 mm/annum or 35 mm/ka."⁸  Their Table 7 compiles ten earlier measurements from Yorkshire, Derbyshire, Slovenia and northern Norway.  The full range is 0.003 to 0.100 millimetres per year.⁸

Now price the top of the Redwall Limestone.  The valleys cut into it "average about 300 m in width and 75 m in depth" and reach 122 metres, and solution caves inside the limestone are filled with Late Mississippian sediment.²⁰  At 0.100 mm/yr, the fastest rate measured anywhere, 75 metres takes 750,000 years.  At the median of the table, about a million and a half.  A Flood year at the fastest rate removes 0.1 millimetre.  Grant a thousandfold speed-up for hot, acidic Flood water and the year buys 10 centimetres.

Two honest caveats.  Billingsley and Beus credit the valleys to stream erosion after regional uplift, so the dissolution rate prices the caves and the plateau between the valleys rather than the valley floors.  And most karst dissolution is focused down joints and at sinkholes rather than spread evenly, so a single doline deepens faster than the surface average.  Treat the number as a floor, good to an order of magnitude.  The floor is still five to seven orders of magnitude away from one year.

The young-earth side has two answers, and they contradict each other.  Emil Silvestru argues that pre-Cenozoic paleokarst is not karst, and supplies a mechanism:  "It is likely that the rapidly accumulating Flood sediments periodically emerged.  The resultant degassing and dewatering of waterlogged sediments, along with the associated chemical changes, would have developed a wide range of negative relief in a matter of days."²¹  John Reed and Michael Oard apply that to the Redwall, adding that the features "can be produced by the Flood, by hydrothermal flows or gas seeps."²²  Steven Austin says the opposite.  The solution is real, but it happened after the whole column was in place, because the Grand Canyon breccia pipes carry clasts of Coconino Sandstone and Kaibab Limestone.  "The evidence for solution occurring in the time interval between Redwall and Supai remains doubtful."¹⁶

One observation answers both.  The caves and valleys at the top of the Redwall are filled with the Surprise Canyon Formation.  It carries Lepidodendron logs and twenty-two spore species "indicative of Upper Mississippian strata," and it is "present only in paleovalleys and karst caves eroded into top half of Redwall Limestone."²⁰⁴  Degassing relief in soft sediment does not make caves that later fill with Mississippian plant debris.  A cave dissolved after the Kaibab cannot be filled with it either.  The fill dates the dissolution.  Reed and Oard's reply on this point is that "time is only in the minds of uniformitarians;  plant matter was present in floodwaters."²²  That answers where the plants came from.  It does not answer why they are only in the holes.

Glaucony is a stopwatch for a starved seafloor

Roth's eighth question asks whether a submerged gap would be protected from erosion.  He is right that it would.  A starved seafloor does something else instead.  It grows minerals in place, slowly, and the chemistry records how long.

Glauconite forms on the seafloor where almost no sediment is arriving.  A grain matures through stages, and its potassium content rises as it goes.  Working on Weddell Sea cores, Adrián López-Quirós and colleagues state the timing:  "The time required to produce an autochthonous evolved glaucony grain is considered to be higher than" 10^4-10ars for ancient glaucony-bearing records," and their figure 7 breaks the process into steps, with the Fe(III)-smectite to glauconite reaction alone running "~10^4-106 years."¹⁰

The flat-gap argument has never engaged this reply.  Dunbar and Rodgers listed glauconite concentrates among the criteria in 1957 and flagged them as evidence of "extremely slow deposition."¹  Roth 1988 never uses the word.  Roth 2009's weathering question does not mention it.⁹  Walker's paleosol and hardground articles do not address it.  A mature, potassium-rich glaucony grain on a paraconformity is a claim about elapsed time.  It owes nothing to biostratigraphy, nothing to radiometric dating, and nothing to the shape of the surface.

Here is the gap on this side of the argument.  The base of the Chattanooga Shale and the overlying Maury Formation both carry glauconite and phosphate, and nobody has published potassium-oxide maturity data on them.

Hardgrounds, and the best exchange on record

A hardground is a patch of seafloor that turned to rock while it was still the seafloor.  Animals bored into it, others cemented themselves on top, and later sediment buried the whole thing.  In 2006 Mark Wilson, who works on Palaeozoic hardgrounds at the College of Wooster, put the case to Creation Ministries International.  Tas Walker answered him at length, and John Woodmorappe published a technical treatment the same year.²³

Wilson's case concerns sequence.  It says nothing about rate.  "The encrusters had to land as larvae on the hardground after it was lithified, and then they developed into adults."¹¹  "The borings often cut into other borings from previous generations, many of which are filled with cemented sediment."¹¹  And the detail that is hard to explain any other way:  a boring bivalve drills a hole that widens with depth as the animal grows, so "The result is a hole with a bivalve in it far too large to get out the restricted opening."¹¹

Walker's reply is a physical model.  Hardgrounds are pre-Flood seafloor debris swept into lag deposits.  Cementation happened after burial as the sediment pile dewatered.  The borings are compaction moulds.  "As the sediment was dewatered, the hardground squeezed the original hole, a mould rather that a burrow, and pushed the organism out like a stone from a squeezed plum."¹¹  He adds a field observation worth taking seriously:  "the burrows and holes in the hardgrounds have not disturbed the substrate much."¹¹

The model makes a prediction, and that is its strength.  A boring is cut into rock that is already hard, so its wall should truncate grains, ooids and any cement already present.  A compaction mould can only deflect grains around itself.  That looks like a thin-section test.  Woodmorappe has published the reason it is not a clean one:  a clean cut through a hard grain may show only that "the hardness of the object itself is a sufficient explanation for this observation," and need not show that the whole layer was lithified.²³  He is right.  So the test has to be narrowed to something a firm matrix cannot supply:  cement bridging two grains, cut through by the boring wall.  No petrography of that kind was checked for this article.  Walker's model has a second problem he does not address.  It needs the encrusting animals to have grown on a hardground that, on his own account, did not become hard until after burial.

Wilson's concession must not be buried either:  "Forming a hardground probably takes years, but we do not yet have any direct way to estimate the duration of the process."¹¹  On hardgrounds the mainstream has a strong argument about sequence and no clock at all.

Part 6: What is on the contacts when somebody walks them

The Grand Canyon supplies most of the photographs in the flat-gap literature.  The canyon's own contacts are treated in detail in The Grand Canyon.  Three results from there bear on this argument, and they are summarised rather than repeated.

The flagship flat gap is channelled.  The Devonian Temple Butte Formation fills channels cut into the Cambrian Muav "as much as 100 ft (30 m) deep in eastern half of map area," mapped along a 28-mile stretch of canyon wall.⁴⁵  Even the base of the Redwall carries "erosion channels of low relief about 5-10 ft (2-3 m)."⁴  Snelling concedes the Temple Butte channels in his own article.  He writes that "there is often good evidence of erosion, such as gullies and stream channels at the boundary between the Temple Butte and the Muav," then rests his case on the one stretch of the North Kaibab Trail where the Devonian is absent.⁶  Austin concedes the Redwall too:  "In places, a slight degree of relief exists on top of the Redwall, and broad channels, as much as 200 feet deep occur."¹⁶  A photograph taken from a rim 1,600 metres above the river does not resolve a 30-metre channel.

The most-cited flat surface in the canyon is a buried landscape.  The top of the Redwall is the contact Dunbar and Rodgers picked as their Grand Canyon example of a paraconformity in 1957, before the Surprise Canyon Formation was named.¹  Fifty years of mapping turned their paraconformity into a valley system with trees in it.  The measurements are in Part 5.  The creationist replies are in the parent article.

The Hermit-Coconino sun-cracks are contested, and the creationist reply is real.  David White described cracks in the top of the Hermit Shale "reaching a maximum width of over 15 inches and extending downward to depths of 25 feet or more," filled with the first Coconino sand, with "infallen pieces of red shale" caught inside the fill.²⁴  Austin reads them as clastic dikes.¹⁶  John Whitmore, on his own fieldwork, reads them as sand injectites, because "the Hermit lacks the appropriate clays for them to be desiccation cracks."²⁵  That exchange is set out in the parent article.

Part 7: The missing time is getting bigger, and one side is quoting the other backwards

The most-cited mainstream paper in recent creationist writing on this topic is Andrew Miall's 2014 study of the Mesaverde Group in the Book Cliffs of Utah.  Miall took the group's duration from ammonite zones, 81.86 to 77 million years.  He compared the sediment actually present against how fast modern rivers, deltas and shorelines deposit.  His conclusion:  "less than 10% of the 4.86 m.y. elapsed time represented by the group can be accounted for."²⁶  On the surfaces themselves he concedes as bluntly as any creationist has:  the intervals between sheets "have not left any identifiable signature, such as mature paleosoils, or evidence or early cementation … or of deep erosion."²⁶

Oard's 2016 note turns that into:  "Miall further notes these obscure gaps are typical of strata worldwide and that sedimentation provides a record of only 10% of the time, while 90% is not represented by any strata at all," and concludes that "the claimed millions of years are imaginary."¹⁵¹⁵

Compare the two.  Miall's 4.86 million years is an input, fixed by ammonite zones and kept.  His result is that the sediment occupies less of that span than anyone assumed.  That means the gaps are longer than previously thought.  Oard reads the same 10% as evidence that the span never happened.  The paper that says "most of this time is not represented by rock" is being cited for "this time is imaginary."

Sedimentologists are arguing among themselves about how to talk about this.  Angela Hessler argued in 2023 that the language of incompleteness imports a false expectation, and that "Stratigraphy is not missing anything."²⁷  Miall answered in 2024:  "Incompleteness is a fact, but it is measurable, quantifiable, and explainable."²⁸  Both positions are nearly seven orders of magnitude from a Flood year:  4.86 million years against one.  Hessler argues that a rock record full of gaps is a normal one.  Miall argues that the gaps are real and countable.  Both accept the 4.86 million years.

Where the young-earth camp stands

Roth and Oard argue that these contacts show little or no erosion, and so little or no time.  The Institute for Creation Research's flagship stratigraphic programme agrees with them.  Clarey and Werner map six megasequences across five continents from roughly 3,000 columns.  They define the unit this way:  each is "a discrete package of sedimentary rock bounded top and bottom by inter-regional unconformity surfaces across the North American craton," and the packages "can be recognized by their bounding erosional surfaces and sudden changes in rock type, independent of fossil content."¹²  That sentence says how a boundary is found in a column.  It does not say how much rock was removed there.  When Clarey turns to that question he sides with Roth:  "the consistent internal stratigraphy of each megasequence testifies against significant erosion," because "Vast erosion in between each megasequence cycle would have likely destroyed this systematic signature in many locations, if not totally."¹²  He reads the boundary at the end of the Cretaceous the same way:  "in most places, it resembles a disconformity, where the sedimentary beds are parallel above and below the boundary surface, with little indication of any missing sediment or massive erosion."¹²  The heavy erosion his model does call for sits at two other surfaces.  One is the Great Unconformity beneath the first package, which has its own article.  The other is the top of the last, where he points to "the planation surfaces like those that formed at the top of the Tejas."¹²

On the flagship contact, Clarey and Werner make Roth's argument in Roth's terms.  The Tippecanoe package is absent between the Muav and the Redwall, and they read that as "primarily due to non-deposition, with only minimal erosion marking the boundary between megasequences."¹³  The two limestones "may have been deposited one upon the other in a matter of days or weeks with little erosion in between.  This is evidenced by the near planar top of the Muav and its fairly consistent thickness."¹³

So the young-earth camp is not split on these surfaces, and Clarey's version is the more testable one.  Non-deposition with little erosion is also how the mainstream reads a starved, stable platform, so the two readings of the Muav-Redwall contact differ on one thing:  how long nothing arrived.  Clarey's calendar puts a number on it.  The Sauk, Tippecanoe and Kaskaskia packages "were deposited in the first 40 Days of the Flood,"¹² so on his model the Muav-Redwall break lasted days or weeks.  A flat top and an even thickness cannot tell weeks from 100 million years, because both readings predict them.  Only a clock can.  None of the clocks in Part 5 has been read on this contact, and that is a gap in the case against Clarey as much as in his own.  The nearest is higher on the same cliff.  Clarey places the Redwall at the start of the Kaskaskia,¹³ so on his calendar the caves and karst at its top were dissolved and filled inside the Flood year.¹²  Part 5 prices that dissolution at 750,000 years at the fastest rate ever measured, as a floor good to an order of magnitude.⁸  The Temple Butte channels that Part 6 describes, cut up to 30 metres into the Muav, also show that the top of the Muav is not planar everywhere.  But 30 metres is nothing like the wholesale stripping Clarey is arguing against, and the channels do not by themselves say how long the break lasted.

John Reed and Michael Oard put it the same way in 2018:  "Paraconformities, if real, would represent short breaks in continuous sedimentation, or possibly short time intervals between successive tsunami-like wave sets."²²  That concedes the surfaces are real breaks and disputes only their length, which is Clarey's position too.

Whitmore's 2019 position is different, and it is the strongest creationist answer here.  He does not claim the hiatus is unreal.  He claims the flatness tells you about the environment:  "While flat surfaces are uncommon in terrestrial settings, they are quite common on the seafloor.  So, it is not surprising that we see flat contacts between marine layers, which are most of the rocks in Grand Canyon, even from a conventional perspective.  There is still missing time in between many of these layers from a conventional view, sometimes in excess of 100 Ma, but at least that is easier to explain in a marine setting."²⁵  That is most of the mainstream answer, published in a creationist journal.  It is also a concession that the flat-gap argument as Roth states it does not work.

Objections and responses

"The surfaces are flat over hundreds of kilometres.  Nothing plane-cuts a continent and then stops."

Something does:  a shallow sea over a stable craton.  Conant and Swanson's answer, on a surface that bevels twenty-three formations, is "a time of unusual crustal stability until a wide area was reduced to an almost featureless plain barely above sea level."³  Wave base and sea level are a flatness machine, and they work slowly.  The Flood alternative has to plane the same surface in weeks and then deposit on it without disturbing the bevel.  The leading young-earth stratigraphic model does not offer that planation.  Clarey and Werner argue the opposite of it at the megasequence boundaries:  "the consistent internal stratigraphy of each megasequence testifies against significant erosion."¹²  At the Muav-Redwall contact they have the two limestones laid down "in a matter of days or weeks with little erosion in between."¹³  A model built on little erosion at its boundaries still has to account for a flat surface that does cut across twenty-three formations, and Part 4 found no young-earth paper that engages the Chattanooga map.

"There is no soil, no channel and no weathering at most contacts.  You have named the handful of exceptions."

Partly fair.  Roth's own concession, that deep erosion is present "rarely," is correct as a description of what is visible.²  But "no channel" is not the same as "no time," and the difference is testable by a subcrop map, a glaucony maturity stage, or a cave with datable fill.³²⁰¹⁰  Three of the eight criteria in the 1957 list have never been applied to a contact by anyone arguing the flat-gap case.

"Mainstream geologists themselves admit they cannot see the time."

They do, and this article prints their words:  Dunbar and Rodgers on subjective recognition, McKee on an obscure flat surface, Miall on surfaces with no identifiable signature, Wilson on having no direct clock for hardgrounds, Newell on not having a simple solution.^15261114  None of them says the time did not pass.  Miall's point runs the other way:  the gaps are longer than the field assumed.

"The missing-biozone criterion assumes evolution, so it cannot be used to prove long ages."

This is Austin's reply, and it is the best one available.¹⁶  It does not hold.  The criterion assumes two things, and neither is evolution.  It assumes the zones occur in a fixed order, which is an observation repeated across continents and boreholes.  And it assumes the same zones are missing across a platform while present and thick in the adjoining basin, which is a mappable prediction that a local scour cannot make.  The Nevada Test Site column is that prediction coming true.¹⁷

"Dating the karst by what fills it is circular."

It would be if the fill were dated by its position.  It is not.  The order is physical:  dissolve, then fill.  The fill is dated by its own spores and logs.²⁰

"This is an argument from ignorance in both directions.  Nobody can see time in a rock."

Fair as stated.  It is the reason this article leans on rates rather than appearances.  Limestone dissolution is measured on datable engineering works;⁸ outcrop erosion comes from a compilation of 1,599 isotope measurements;⁷ glaucony maturation is mineral kinetics.¹⁰  A rate times a measured thickness is a duration.  That appeals neither to what a surface looks like nor to consensus.

"You conceded 540 metres.  Where did that erosion go?"

The 540 metres is what the outcrop rate predicts for a surface exposed to the air for the whole 100 million years.  The geology says it was not.  A platform at or near sea level under a shallow sea erodes subaerially for only part of the interval, and that is the only part the beryllium-10 rate applies to.  This is the weakest joint in the mainstream answer, because "it was submerged most of the time" is hard to test directly.  The independent check is the condensation minerals, and it has not been made.

"Snelling's locality has never been re-surveyed.  Why should either side be believed?"

It has not, and that is a real hole.  Snelling's whole claim rests on a 1986 field survey by Waisgerber, Howe and Williams in the Creation Research Society Quarterly.  It contradicts McKee and Gutschick's published diagram, which shows an irregular wavy surface with one to two feet of channelling relief.⁶  Neither document was obtained for this article.  Two field surveys of one 200-metre stretch of trail disagree, and nobody else has gone to look.  The argument in this article does not rest on that locality, and it should not.

Issues and unanswered questions

• Two sources that matter were not obtained. Peter Sadler's 1981 completeness paper is paywalled, and nothing here depends on his own wording. Stanley Beus's 1990 Temple Butte chapter, which Roth quotes without a page number, is in no scan. McKee's public-domain professional paper says the same thing and is used instead.
• Roth's Van Andel and Beus quotations are not printed here. Roth's two printings of the Van Andel sentence differ from each other, and the Beus quotation cannot be checked against the chapter at all. Both are in quarantine with what was tried.
• The glaucony measurement that would close a named case has not been made. Conant and Swanson report glauconite and phosphate at the base of the Chattanooga Shale and in the Maury Formation. Nobody has published potassium-oxide maturity stages on those grains. That is one microprobe session against a type paraconformity.
• No thin section has been checked here. Woodmorappe's objection to the petrographic test of hardground borings has not been answered from the bioerosion literature, and that search was not run for this article.
• Nobody has re-surveyed the North Kaibab contact. Two field reports of one stretch of trail disagree, both are decades old, and neither was read in the original for this article.

Verdict

The observation is sound.  The inference is not.  Paraconformities are real, the discipline named them, and at many contacts there is nothing at all to see.  Roth states that honestly and Whitmore states it better.

But "nothing to see" was tested along one dimension.  The evidence that decides these surfaces is invisible at an outcrop:  a subcrop map that shows twenty-three formations truncated under one bedding plane;  a dissolution rate that prices 75 metres of karst at a minimum of 750,000 years;  a mineral clock that needs 10^5 years on a seafloor where nothing is happening;  a biozone missing across a whole craton and a kilometre thick in the basin next door.  Every one of those is a measurement.  Two have young-earth answers:  Silvestru's degassing relief for the karst, and Woodmorappe's objection to the boring petrography.  Neither supplies a rate, and neither explains why the Redwall fill sits only in the holes.  The Chattanooga bevel has no answer at all.  The one young-earth article on that shale describes the outcrop and never opens the map.

The flat gaps are not evidence that no time passed.  They are evidence that most of geological time deposited nothing.  That is what a starved, stable, shallow platform does, and it is what the founding textbook of stratigraphy said in 1957.

Sources