Coordinate Cosmologies: Dennis and Tenev

Summary

Two young-earth cosmologies published in 2018 answer the distant starlight problem the same way.  Neither changes the physics.  Both change the labels.

Phillip Dennis, a physics PhD who worked on gauge field theories and general relativity, takes the standard big bang spacetime, cuts away part of its past, and calls the cut surface the first day of creation.¹¹  Tichomir Tenev, John Baumgardner and Mark Horstemeyer keep flat spacetime intact and define a new time coordinate on it, the elapsed time since creation at each place, which they call the Creation Time Coordinate.²  Both proposals are careful.  Both are consistent with relativity.  Both were presented at the same conference, in the same volume, a few dozen pages apart.  And both, by their authors' own statements, predict exactly what standard cosmology predicts.

That last sentence is the point of this article.

• Dennis says his model is a piece of the standard model. His words: the solution "is an open subset of the FLRW space-time so that all differential properties and analysis that applies to FLRW also applies to our solution."¹ FLRW is the standard big bang spacetime. An open subset of it has the same physics, the same light propagation, the same redshifts. He states the consequence himself: "the predictions of red-shifts are identical to those of the FLRW cosmology."¹
• Tenev and colleagues say the same about their convention. Switching between the Einstein convention, Lisle's anisotropic convention and their own "alters the coordinate axes, changes how coordinates are assigned to events, and consequently the events' perceived ordering, but does not change the relationship among the events in spacetime."² Their appendix is blunter: the choice of synchrony convention "has no physical significance at all."²
• Four young-earth models, four authors, one result. Dennis on his own spacetime, Jason Lisle on the anisotropic synchrony convention, John Hartnett on his bounded Carmelian universe, and Samuel Conner and Don Page on Russell Humphreys' bounded universe all report the same finding: the model is observationally identical to the standard one.¹³⁴⁵ None of the four has been refuted by an observation. None of the four has ever predicted one.
• The three "different" surfaces are one family. Lisle's surfaces of simultaneity are past light cones. Tenev's are hyperboloids just inside the light cone. Dennis's creation surface is a hyperboloid in the same family, with a free parameter b that becomes the light cone as b goes to zero.¹ Tenev and colleagues notice the convergence and say so: their surfaces and Lisle's "converge asymptotically with each other at large distances," and Lisle's convention applied to Earth "is essentially equivalent to our definition of the CTC."² Three proposals, three names, one geometry.
• Tenev and colleagues admit that the convention does no work. This is the most useful sentence in either paper: "it is the initial conditions that make the solution possible and not the convention."² The initial condition is that God created every star just outside the past light cone of Earth's Day Four. That is a claim about where things were put, not about how time is labelled, and it is unfalsifiable for a different reason: it is a statement about a single miraculous act with no independent evidence.
• Dennis is the sharpest published critic of Lisle's model and has never explained why his own escapes the same objection. He argues that under Lisle's convention "the purported speed of light is not a physical speed" but "a coordinate rate," and that the convention is therefore "a vacuous coordinate transformation."⁶⁷ His own 2018 figure caption reads: "the global speed of incoming light approaches infinity as distance increases … Since the model is a solution of the EFE, the local speed of light is constant."¹ That is precisely the distinction Lisle draws. Across thirty pages of his second attack on Lisle, Dennis mentions his own cosmology twice, both times on the first page, and never brings it into the argument.⁷
• The Dennis and Lisle exchange now runs to fourteen publications and is finished. Two Answers Research Journal papers by Dennis in 2024, Lisle's July 2024 reply, and an eleven-part series on Lisle's site running from September 2025 to 28 August 2026, when Lisle closed it.⁸ Parts five through eight are about the metaphysics of time, not physics. Each man calls the other's position heretical.⁷⁸
• One earlier grade in this project needs correcting. The article on the anisotropic synchrony convention records Dennis's general-relativity objection as unanswered. Lisle answered it on 22 May 2026, in part nine, and the answer is half good. He points out that changing the time slicing is routine in curved spacetime, citing his own peer-reviewed paper on the Gullstrand-Painlevé form of the Schwarzschild metric.⁹¹⁰ What he does not answer is Dennis's actual point, which is not that other slicings exist but that in a matter-filled universe one of them is picked out by the matter.
• What a coordinate change cannot do is shorten a history. Proper time, the time a clock carried along with an object actually records, is the same number in every coordinate system. Relabelling t does not un-cool a white dwarf or un-expand a supernova remnant. The cleanest single case is the ring around SN 1987A: 0.67 light years in radius, expanding at 10 kilometres per second, so it was thrown off about 20,000 years before the star exploded.¹¹ Under either of these models the star was created about 6,000 years ago. The ring has to have been created already in flight. The SN 1987A article works this through.
• The honest verdict. These are the two least embarrassing young-earth cosmologies in print. They use real general relativity, they do not need new physics, they do not require the speed of light to have changed, and their authors state their limits in the papers themselves. A young-earth reader looking for a model that will not fall apart under a physicist's questions should read these two rather than the popular literature. What they buy is the removal of one argument. What they cost is everything a model is normally for: a cosmology built to reproduce every observation predicts none of them.

What a time coordinate is

A coordinate is a label.  Nothing more.

Give every event in the universe four numbers, three to say where and one to say when, and you have a coordinate system.  The numbers are a bookkeeping device.  They are not the events.  Renaming a street does not move the houses on it.

The t in the equations of physics is one of those labels.  It answers the question "what time do you call that distant event?"  For events at the same place as you, the answer is easy:  you look at your watch.  For events far away, the answer requires a rule, because the light carrying news of the event took time to arrive, and you have to decide how much.  That decision is called a synchrony convention.  Different conventions assign different t values to the same distant event.  All of them agree about your watch.

Here is the freedom that matters, stated once and precisely.

Relativity fixes the round-trip speed of light in vacuum.  Send a pulse to a mirror and back, time it on one clock, divide the total path by the total time, and you always get 299,792,458 metres per second.  That is measured to enormous precision.  Relativity does not fix how that round trip splits into an outbound half and an inbound half, because measuring one half alone would require two clocks in different places already agreeing on the time, and setting those clocks requires knowing the one-way speed.  The circle has no exit.  Einstein said so himself, calling the equal split "a stipulation which I can make of my own freewill."¹²¹²  The anisotropic synchrony convention article works through the fifty years of philosophical argument this has generated.  Both papers examined here take that freedom as their starting point, and both are entitled to.

Now the limits, which are just as firm.

A coordinate change cannot alter the outcome of any experiment.  The reason is structural, not empirical.  Physical laws in relativity are written as relations between tensors, which are objects defined so that their content survives relabelling.  Change the labels and every equation changes shape while every measurable number stays put.  This is why nobody in physics argues about which coordinates are "true."  They argue about which are convenient.

A coordinate change cannot alter an invariant.  Some quantities carry the same value in every coordinate system by construction.  Two of them run this whole article.  The spacetime interval between two events is one.  Proper time, the elapsed time actually recorded by a clock carried along an object's own path, is the other.  Proper time is what ages things.  A star's fusion history, a white dwarf's cooling curve, a supernova remnant's expansion, and the decay of a radioactive nucleus all run on proper time.  No relabelling of t touches any of them.

A coordinate change can move where the "now" line is drawn.  This is the freedom both models use, and it is real.  Take the set of all events you are willing to call "at the same time as this moment."  In the ordinary Einstein convention that set is a flat slice through spacetime.  Draw it differently and you get a curved slice.  Relativity permits many such slices.  Which events count as "now, far away" has no observer-independent answer, and cosmologists exploit this all the time:  they routinely slice spacetime by past light cones and call the result observational coordinates.

So the honest summary of the freedom is this.  You may choose where to draw "now."  You may not choose how long anything has been running.

Both of the models below take the first freedom to its limit.  Neither can take the second, and neither claims to.

Dennis: cutting a piece out of the big bang

Who is making the argument

Phillip W. Dennis holds a physics PhD from the University of Missouri.  His stated fields are gauge field theories with spontaneous symmetry breaking, quantum field theory and general relativity;  he has three patents, awards from NASA, TRW and Litton, and membership in the American Physical Society.¹  He states his own position at the top of his 2024 paper:  "I am a young earth creationist who is thoroughly committed to a literal seven days of an actual Creation Week."⁷

He presented "Consistent Young Earth Relativistic Cosmology" at the Eighth International Conference on Creationism in Pittsburgh in 2018.  It runs twenty-two pages with two appendices.¹  The paper was hard to obtain for years, because Cedarville University migrated the conference proceedings to a new host that refuses browsers, and this project only retrieved it in August 2026.¹

What he does, in plain language

Start with the standard big bang spacetime.  Cosmologists call it FLRW, after Friedmann, Lemaître, Robertson and Walker.  It is a four-dimensional geometry:  a stack of expanding three-dimensional spaces, one for each moment of "cosmic time," running from a beginning about 13.8 billion years ago to now.

Dennis does not change that geometry.  He deletes part of it.

He draws a curved surface through the stack.  Near Earth the surface sits high, only about 6,000 years below the present.  Far away it dives down toward the beginning, so that a galaxy ten billion light years off has its slice of the surface ten billion years back.  Everything below that surface, he says, never existed.  The surface is the first day of creation.

The trick is in the shape.  The surface is drawn so that it hugs the past light cone of Earth's creation without ever touching it.  A light ray leaving a distant galaxy the instant that galaxy is created therefore arrives at Earth almost immediately in Earth's own time, because the galaxy was created almost exactly as far back as the light needs to travel.

His own summary of the construction:

"We build the model from the closed FLRW solution by selecting a spacelike hyperboloidal surface as the initial surface at the beginning of the first day of creation. … The resulting space-time is a subset of the usual FLRW space-time and thus preserves the FLRW causal structure and the observational predictions such as the Hubble law."¹

And on the deletion, with no hedging:

"With this choice we are necessarily taking a subset of the usual FLRW, since we are discarding regions of the FLRW manifold … Viewed as a cross section of the FLRW cosmology the portion of the FLRW below the surface labeled 'Day 1' did not exist."¹

The mathematics, kept small

FLRW spacetime can be written in a form where light travels on straight 45-degree lines, using a rescaled clock called conformal time η and a comoving radial distance χ:

ds² = a²(η) [ −dη² + dχ² + f_K(χ) dΩ² ]

A radial light ray coming in toward χ = 0 satisfies dη = dχ, so it moves one unit of η per unit of χ.  Dennis's creation surface is

η = τ − √(χ² + b²)

where b is a single free constant.¹  Set b = 0 and the square root becomes χ, so the surface is η = τ − χ, which is exactly the incoming light ray:  the past light cone.  Any larger b bends the surface a little away from that cone.

b has a plain meaning, and Dennis gives it:  it is "the time it takes light from objects at infinity to reach the earth."¹  He then relabels time by defining a new coordinate τ = η + √(χ² + b²), which is constant along the creation surface, and substitutes a variable ρ through χ = b sinh ρ.  The metric that comes out is a rewriting of FLRW in these new labels, and the elapsed proper time of a galaxy sitting at ρ is the integral of the scale factor along its own path.

Two consequences follow, and Dennis draws both.

Every observation is unchanged.  He is explicit and repeats it three times.  The spacetime "is a subset of the usual FLRW space-time and thus preserves the FLRW causal structure and the observational predictions such as the Hubble law";¹ "all observable physics will be in line with the equations of GR and the implications of the FLRW metric … since the solution is in fact a section of the FLRW manifold.  All tensorial equations and invariant quantities are the same.  In particular, the predictions of red-shifts are identical to those of the FLRW cosmology";¹ and finally, in the conclusion, the model "reproduces the observational consequences of both the FLRW space-time ('big bang') and any modifications such as inhomogeneities."¹

The whole universe becomes visible after a time b.  He computes the particle horizon and notes that it runs to infinity as the elapsed time approaches b:  "This shows that the entire universe is observable after the passage of time b."¹  In his own plotted example he uses b equal to 1,000 years.

The part that is easy to miss

Look at the caption Dennis wrote under his own light-ray diagram:

"Note that the global speed of incoming light approaches infinity as distance increases.  Conversely the global speed of outgoing light decreases with distance.  Since the model is a solution of the EFE, the local speed of light is constant in accordance with the principle of relativity."¹

Read that beside his 2024 attack on Lisle, in which the fatal defect of the anisotropic synchrony convention is that "the purported speed of light is not a physical speed.  The anisotropic speed of ASC is a coordinate rate not a physical speed and the word speed in ASC is used in an equivocation."⁶

Dennis's own model has an inbound coordinate light speed that goes to infinity with distance and a local physical light speed that stays at c.  That is the same two-level structure, drawn by the same hand, six years earlier.  Nothing in the 2018 paper is wrong about it.  The problem is that the 2024 papers never mention it.

The creation week, and where the miracle sits

Dennis's construction gets distant light to Earth within a time b.  It does not by itself get that light to Earth on Day Four.  With b at 1,000 years, Adam saw nothing beyond a thousand light years.  Dennis knows this, and his answer is a second, separate move.

"However, there is nothing to preclude God from advancing the remote regions more rapidly thereby yielding a non-null hyper-surface.  That concept is consistent with the biblical account."¹

His figure of that variant is captioned "Possible development of the past spacelike surfaces during the creation week in which time is miraculously advanced more rapidly at distant locations," and it ends with:  "it should be noted that, as drawn, the entire cosmos (in fact, objects at possibly infinite distances and look back times) is visible at the earth on Day 7."¹

Notice what has happened.  Distant regions run through more time than Earth does during the same week.  That is time dilation, obtained by decree rather than by gravity.  It is the same shape of answer as Russell Humphreys' white-hole cosmology, which Dennis lists among the failed attempts on his second page.¹  He does not quantify how much extra time the remote regions get, and the figure is explicitly "not to scale."  This is the weakest joint in the paper, and a reader should see it clearly:  the version without a miracle does not deliver starlight on Day Four, and the version with one has not been costed.

What Dennis takes to be the objection, and his answer to it

The paper's last substantive claim is a challenge:

"The model only differs from conventional solutions in that it uses an initial condition ('creation hypersurface') that supports a young earth.  There can be no unprejudiced objection to this solution.  A prejudiced objection would rest only upon the assumption of requiring a naturalistic metaphysics requiring, in some sense, a synchronized (and simultaneous) explosion of all matter from a white hole by extrapolating a current value of the metric backward in a presumed preferred time to a putative initial singularity.  Such an assumption is not dictated by GR."¹

The technical half of that is right.  General relativity does not tell you which spacelike surface to call the beginning.  Choosing the standard one, on which the big bang happens everywhere at once, is a modelling decision motivated by the observed uniformity of the microwave background and the matter distribution, not a theorem.  Dennis is correct that a different initial surface is mathematically admissible.

The rhetorical half is not right, and it should be said plainly.  "There can be no unprejudiced objection" is too strong, and Dennis supplies the objection himself two pages earlier.  He writes:

"Relativistic principles do not distinguish any preferred initial geometry or any preferred cosmological simultaneity surface."¹

If no surface is preferred, then his hyperboloid is not preferred either.  The model is not selected by evidence;  it is selected because it "supports a young earth," which is his own phrase.  A cosmology with a freely choosable initial surface, tuned to a desired conclusion and predicting nothing else, is not a model that evidence can support.  It is a model evidence cannot touch.  Dennis has produced a proof that a young-earth reading is compatible with general relativity.  That is a real result and it is worth having.  It is not a cosmology in the sense that FLRW is one.

Tenev, Baumgardner and Horstemeyer: creation time coordinates

The proposal

Tichomir Tenev was then a doctoral student in computational engineering at Mississippi State University with two MIT degrees;  John Baumgardner has a geophysics PhD from UCLA and spent most of his career at Los Alamos, and did the original work behind catastrophic plate tectonics;  Mark Horstemeyer is a mechanical engineering professor at Mississippi State and a fellow of four societies including the AAAS.²  Their paper appeared in the same 2018 conference volume as Dennis's, sixty-eight pages later.²

Their proposal has two halves, and they are careful to keep them apart.

The initial condition.  God placed the creation event of every star at a spacetime point just outside the past light cone of Earth's Day Four and inside the past light cone of Earth's Day Five, with every stellar creation event causally independent of every other.  Geometrically, the stars were created along a hyperboloid "whose slope is everywhere shallower than the slope of a light cone."²²  Light leaving a star at its creation therefore reaches Earth during Day Four, however far away the star is.

The convention.  The Creation Time Coordinate of any event is "the elapsed time between that event and Creation Day Four at the event's location, plus three days," which is the same as the elapsed time since Genesis 1:1.  Two events are simultaneous when their CTCs are equal.²

Their claim is that God's numbering of the days in Genesis 1 is this convention.  Declaring the stars made on Day Four is, on their reading, a divine stipulation that all those creation events are to be counted as simultaneous with each other and with Earth's Day Four.

Why this is a real improvement on Lisle

Lisle's convention is defined relative to an observer.  Anyone, anywhere, may adopt it, and each gets a different set of simultaneity surfaces.  Tenev and colleagues regard this as the weakness that has kept the model from being accepted:

"The ASC is an observer-specific and hence subjective definition of simultaneity, while the CTC-based synchrony convention is a divinely-prescribed and hence objective definition of simultaneity."²

Their CTC is tied to one place and one event, so there is exactly one of it.  They also identify a real gap in Lisle's presentation:

"In its essence, our solution is a reformulation of Lisle's solution (Newton 2001; Lisle 2010), but ours spells out clearly the required initial conditions, without which Lisle's solution is at best ambiguous if not incomplete."²

The point is well taken.  A synchrony convention on its own says nothing about where anything was put.  Adopt Lisle's convention with different initial conditions and the distant stars need not be visible from Earth at all.  Tenev and colleagues say this in as many words:  "using different initial conditions is admissible within Lisle's ASC model; hence, distant starlight might not arrive at Earth at all during the time of the Earth's existence."²

The sentence that decides the matter

Having separated the two halves, they then say which one does the work.

"The initial conditions, which are independent of the choice of synchrony convention, are fundamentally what enable distant starlight to arrive at Earth on Day Four.  All that is needed is for stellar creation events to be positioned just above the Earth's Day Four past light cone.  The purpose of the synchrony convention is to prescribe an absolute ordering of these events."²

And again, answering the objection that the model defines the problem away:

"As pointed out earlier, however, it is the initial conditions that make the solution possible and not the convention."²

This is the most useful admission in the entire young-earth starlight literature, and it deserves to be stated in both directions.

Against the critic:  the model is not a word trick.  Something substantive is being claimed, namely that every star in the universe was created at a specific spacetime point selected so that its first light would land on Earth on one particular day.

Against the model:  that substantive claim is a bare stipulation about a single act of creation, supported by no evidence and capable of producing none.  The convention was the part that could be defended by physics, and the authors have just told the reader that the convention is not what solves the problem.  What is left carrying the weight is an initial condition with one free function, chosen to give the answer wanted.  It is the same structure as Dennis's freely chosen creation hypersurface.  Both models put the whole explanatory load on a surface drawn by hand.

The asymmetry, which is a genuine prediction of a sort

One consequence is worth noticing because it is unusual and the authors face it squarely.  If starlight reaches Earth instantly by the CTC clock, the reverse does not hold:

"In fact, due to the special initial conditions, a star located a billion light years away from Earth will not receive light, or any other signal from Earth, until its CTC clock strikes two billion years."²

They defend this from Genesis 1:15 and 1:28:  the stars were appointed to give light on the Earth, and man was not given dominion over them.  Whatever one makes of the exegesis, the asymmetry is not hidden, and it is a real structural feature of the model:  Earth occupies a privileged position in spacetime, not merely in space.  They say so and accept the cost, rejecting the cosmological principle explicitly.²

Their worked example, and why it is the model in miniature

Their illustration is SN 1987A, and it shows exactly what the CTC does and does not achieve.

"In terms of CTCs, one would date the star explosion to be about 4,000 + 1,987 = 5,987 years since creation, assuming creation took place about 4,000 years BC, and the star's creation to be Day Four since the beginning of Creation."²

Note carefully what has changed.  The date 166,000 BC becomes 5,987 years after creation.  Nothing about the supernova has changed.  The neutrinos still arrived three hours before the light.  The ring still lit up over 322 days.  The progenitor was still a blue supergiant.  What changed is the number written next to the event in a table.

This is the model working exactly as designed, and it is also the model's ceiling.  Relabelling the explosion does not relabel the ring.

Three surfaces, one family

Tenev and colleagues include a diagram comparing the three conventions, and its caption gives the geometry cleanly.  The surfaces of simultaneity are, respectively, "hyperplanes, hyper cones, and hyper-hyperboloids" for the Einstein convention, Lisle's convention, and their own.²

Put that beside Dennis's creation surface, which is a hyperboloid with a free parameter that becomes a light cone in the limit, and the picture is plain.  The Einstein convention slices spacetime flat.  Lisle slices it along past light cones.  Tenev and Dennis both slice it along hyperboloids that hug those cones.  These are one family of surfaces with a single dial on them.

The authors notice half of this themselves.  Comparing their surfaces to Lisle's, they write that "although the isochrone hypersurfaces differ slightly, hyper-cones in one case versus hyper-hyperboloids in the other, they converge asymptotically with each other at large distances," and conclude that Lisle's convention as he actually applies it "is essentially equivalent to our definition of the CTC."²  They do not compare their surfaces to Dennis's, and Dennis does not compare his to theirs.  The two papers appear in the same volume and neither cites the other.

Their observational argument, and the problem with it

Tenev and colleagues do not merely propose;  they offer evidence.  Their claim is that the sky looks young, and their main line is the shortage of old supernova remnants.

Their treatment is better than the argument's reputation.  They audit Keith Davies' 1994 numbers and find an arithmetic error in Davies' favour:

"Davies did include that factor in his estimate of the expected number of observable SNRs were our galaxy much older than the SNR maximum lifetime, but he failed to include it for the case of a 7,000-year age.  Including it in that case yields only 126 expected SNRs instead of 268.  That number is notably fewer than the 200 actually observed as of 1994.  On the face of things, that would suggest that either the galaxy is somewhat older than 7,000 years or else the estimated average time between SN explosions is too large.  We suspect that the latter is the more likely explanation."²

Correcting an error against your own side, in print, is the mark of honest work, and it should be credited.

What survives their correction is the comparison of 200 observed Sedov-phase remnants against 2,256 expected if the Galaxy is older than the Sedov lifetime, plus the stronger claim that "there are no observed third (snowplow) phase SNRs, given that they are expected to persist as readily detectible entities for hundreds of thousands of years."²

That second claim had been contradicted the year before by the staff astronomer at Answers in Genesis.  Danny Faulkner's 2017 paper is titled "Are Old Supernova Remnants Really Missing?" and its abstract says:  "reexamination of the data shows that there indeed are old supernova remnants, and that stage three supernova remnants have been identified.  Therefore, this may not be a good argument for recent origin, and I discourage recent creationists from using it."¹³¹³  His table lists seventeen Galactic remnants older than Davies' 7,000-year criterion, five of them above a million years, the oldest at 4.5 million.¹³

Tenev and colleagues do cite Faulkner 2017 and answer him, which is worth stating precisely because this project's parent article previously reported the opposite.  Their reply is on page 87:

"While we do note that some SNRs appear to be older than 6,000 years, such observations may be explainable by yet unknown processes that have caused the SNRs to expand faster than our current models predict.  This is why our argument for young cosmos is based not on the 'reading out' of the apparent age of the SNRs, but on the paucity of SNRs."²

That reply is weaker than it looks, and the weakness is worth naming.  Their whole argument is a comparison between a count of remnants and a count predicted from remnant lifetimes.  Lifetimes come from expansion models.  If those models are wrong by enough to make apparently old remnants young, the expected counts computed from the same models are wrong too, and the shortfall they are pointing at cannot be trusted either.  "Unknown processes" cannot be invoked on one side of a ratio only.  Faulkner's objection is not answered.

The blueshift argument, which is their best original contribution

Buried in their evidence section is a sharp argument that gets less attention than it deserves.  It is aimed at every young-earth cosmology that solves distant starlight with time dilation, including Humphreys' and Hartnett's.

If time ran fast in the distant universe while it ran slowly on Earth, then light emitted during that episode arrives blueshifted, because the source's clock was running fast relative to ours.  Enough dilation to fit billions of years of stellar history into a creation week means enough blueshift to move the light clean out of the visible band:

"If at any period of history, a time dilation event took place that caused time lapse to accelerate in a region of the distant cosmos, then starlight from that region would be blue shifted.  The stronger the time dilation, the greater the blueshift.  To account for the alleged billions of years of stellar history, the blue shift caused by the accelerated time lapse would be so great that it would render the starlight from the affected section of the cosmos invisible."²

Lisle makes the same argument against Humphreys independently.  It is a rare case of a young-earth cosmology paper producing a quantitative objection to a rival young-earth cosmology, and it is a good one.  It applies with full force to the miraculous-time-advance variant of Dennis's own model, which Tenev and colleagues never mention.

The Dennis and Lisle exchange, and where it now stands

The record

The same man who built the FLRW-subset model is also the most persistent published critic of Lisle's convention.  The dispute is now fourteen items long and, as of 28 August 2026, closed by Lisle.

Round one: Lisle is right

Dennis's headline claim is that ASC contradicts itself.  He writes the spacetime interval as

s² = Δx² + Δy² + Δz² − c²Δt²

sets it to zero for a light ray, and gets Δx = ±cΔt in both directions, therefore an isotropic one-way speed.  Lisle's answer is that this diagonal form of the interval is not neutral;  it is the form the interval takes only when the Einstein convention has already been assumed.  The synchrony-general version carries two extra terms that vanish only at ε = ½.  His verdict:  "he assumed that ε = ½ in his attempt to prove that ε = ½.  This is the logical fallacy of a petitio."¹⁴

That is correct, and it should not be softened.  Dennis started from a metric that encodes isotropic one-way propagation and derived isotropic one-way propagation.  Nothing else could have come out.

Round two: Dennis is right about the thing that matters

His conclusion nevertheless survives his failed proof, because Lisle agrees with it:

"The ASC convention is merely a coordinate transform that cannot alter physical quantities.  The conclusion is that ASC does not alter the invariant speed of light or the time that it takes light to reach the earth and therefore does not solve the LTTP."⁶

Lisle says the same thing in his own words repeatedly.  "Changing coordinate systems cannot possibly affect any observational result … This is why it is ultimately impossible to disprove a coordinate system such as ASC or ESC."¹⁵  "No physical observation can refute ASC because it will always make the same observable predictions as ESC."¹⁶¹⁶  "ESC and ASC must always make the same observational predictions since they are merely two different coordinate systems for expressing the same underlying reality."⁹⁹

The two men are not disagreeing about relativity.  They agree about relativity.  They disagree about what kind of problem distant starlight is.  For Dennis it is a physics problem about photons in transit, and a relabelling cannot move a photon.  For Lisle it is an exegetical problem about which convention the Hebrew text uses, and a relabelling is exactly the right shape of answer.  Both readings are coherent.  Neither has been established.

Round three: the empirical arguments fail

Dennis's second paper offers two measurements.  Both fail, and one fails on his own page.

Light-in-flight photography.  He fits Caltech's high-speed movies of a laser pulse crossing a water tank, gets 2.264 × 10⁸ m/s, and concludes that the experiment "empirically establishes the isotropy of the one-way speed-of-light."⁷  A page later he concedes that ASC "logically requires that the transverse speed of light is exactly c in every transverse direction."⁷  A light-in-flight camera photographs a pulse crossing its field of view sideways, and the light it actually records is scattered toward the lens.  Both legs are transverse, where ASC and the Einstein convention give identical answers.  Rotating the apparatus does not help, because ASC's anisotropy is measured from the line of sight to the observer, not from a compass direction.

Slow clock transport.  New Horizons lost less than half a second of proper time in nine and a half years to Pluto, so its clock counts as slowly transported;  compare its elapsed time against ground time and a one-way light travel time of hours falls out.⁷  This is the classic argument, and the classic answer is that "slow" is itself defined relative to a synchrony convention.  Lisle redoes the calculation with the synchrony-general time dilation term and gets four hours and twenty-five minutes of dilation under ASC where the Einstein convention gives 0.41 seconds, which is exactly the light travel time the argument was supposed to establish.¹⁶  Dennis declined to argue the point:  "Undeniably, there has been controversy regarding the slow clock transport method of synchronizing clocks.  I am not going to enter that controversy at this point because it is not essential."⁷  It is the whole question.

Round four: the general-relativity argument, and Lisle's answer to it

Dennis's strongest move is to leave special relativity behind.  The conventionality thesis, he argues, is a result about empty flat spacetime.  The real universe is filled with matter, and in a matter-filled FLRW universe the matter picks out a time:

"All the matter moves along worldlines for which their coordinates are constant, consequently the clocks for each element of matter record the same cosmic time … This is the global absolute cosmic time.  Each value of t specifies a 3D space of simultaneous events.  Thus, this solution has a global time and is a GR counterexample to the conventionalist's claim that there is no such thing as global time or a metaphysical present."⁷

He is right that FLRW spacetimes have a distinguished foliation.  The comoving observers all agree on elapsed cosmic time, and the microwave background picks the rest frame out observationally to within a fraction of a per cent.

This project's ASC article recorded that objection as unanswered.  It is now answered.  Lisle replied on 22 May 2026, in part nine:

"Simultaneity remains both relative and conventional in a non-Minkowski metric.  In fact, I have published on the advantages of using a Gullstrand-Painlevé metric instead of the standard Schwarzschild metric to understand the behavior of black holes (Hamilton & Lisle 2008).  The difference between the two coordinate systems involves a change in the stipulated one-way speed of light and hence the definition of simultaneity."⁹

The citation is real and it is not a creationist venue.  Andrew Hamilton and Jason Lisle published "The river model of black holes" in the American Journal of Physics in 2008, and the paper turns on precisely the point at issue:  Gullstrand "recognized that observables such as the redshift of light from the Sun are unaffected by the choice of coordinates in the Schwarzschild geometry."¹⁰¹⁰  So Dennis's claim that conventionality is confined to special relativity is wrong, and Lisle has a published example to prove it.

Lisle's second point is the better one:

"It is convenient to use a reference frame in which the expansion of space occurs at the same rate in all locations at a given time coordinate in performing our computations.  But this is not a logical necessity. … Yes, the choice to use the Hubble flow as the reference frame (and ESC) is a preference – not a requirement of general relativity."⁹

The answer is half good and the half that is missing should be named.  Lisle shows that other slicings are legitimate.  Dennis never denied that.  His point is that in FLRW one slicing is singled out by the matter, which is not true in the Schwarzschild geometry Lisle's counterexample uses, because Schwarzschild is a vacuum solution with no matter to single anything out.  The question of whether a distinguished foliation is also a metaphysically privileged one is a question about the philosophy of time, and neither man settles it.  Grade the objection as answered but not disposed of.

There is also a strange development here.  Lisle now rejects the FLRW metric outright:  "there is very compelling observational evidence that the FLRW metric is wrong."⁹  Dennis's cosmology is an FLRW spacetime.  The two young-earth physicists have therefore ended up disagreeing not only about how to label time but about the geometry of the universe, and neither has engaged the other's model on that ground.

Round five: the physics stops

Parts five through eight of Lisle's series are not about physics.  They are about whether only the present moment exists.  Dennis is a presentist and Lisle is an eternalist, and each holds that the other's position is unbiblical.

Dennis, on Lisle:  the trouble is "the heretical belief in spacetime as an actual entity and that the past is still there."⁸  Earlier, in the abstract of his second paper, he had promised to expose "the theological implications of Lisle's adherence to the eternalist philosophy of time.  A philosophy, as we will see, which runs afoul of the biblical revelation and, if knowingly held, is a serious heresy."⁷

Lisle, on Dennis:  presentism "is dreadfully antibiblical (it is the philosophy of the heretical view known as 'open theism')," and "if the past is not real, then Jesus did not die on the cross since that was a past event."⁸

There is also an institutional quarrel.  Lisle's opening instalment asks:  "Why were such obvious errors not identified by the peer reviewers of Dennis's paper?  Certainly his paper would never have passed peer review in a professional physics journal.  Could this be why he published it in the Anwers Research Journal?"¹⁵  His footnote is sharper:  "The Answers in Genesis ministry apparently has no access to experts in relativistic physics, which may partly explain why Dennis's erroneous claims were published."¹⁵  By part eleven he records a thaw, noting that "the new leadership of Answers Research Journal invited me to publish my second refutation of Dennis in that journal," an invitation he declined on length grounds.⁸

Two credentialed young-earth physicists, fourteen publications, four of them peer-reviewed by the standards of the venue, and the final four rounds are about the metaphysics of time and each other's competence.  That is what a research field looks like when nobody outside it is checking the work.  Danny Faulkner noticed the same thing about ASC in 2013, complaining that the proposal had not received "more attention, particularly of the negative type."¹⁷¹⁷

The hole in the middle of the exchange

Across fourteen publications, neither man discusses Dennis's own cosmology.

Dennis mentions it twice, both times on the first page of his second paper, as a credential rather than an argument:  "In my 2018 ICC paper (Dennis 2018) I published a general relativity (GR) solution to the LTTP (light travel time problem) consistent with a literal seven days of creation."⁷  It never appears again in thirty pages.  Lisle never mentions it at all, in the July 2024 reply or in any of the eleven instalments, while repeatedly asserting that "the ASC model is currently the only young-universe biblical model of cosmology that (1) is consistent with established physics and (2) has made successful predictions."¹⁵

That claim cannot be right as stated.  Dennis's model is an open subset of an exact solution of the Einstein field equations.  Whatever else is wrong with it, "consistent with established physics" is the one thing it certainly is.  Lisle's real claim must be about the predictions, and there the argument is different and better:  Dennis's model, being FLRW, predicts what FLRW predicts, and Lisle believes JWST has falsified some of that.  But he never says so, because he never engages the model.

Meanwhile Dennis's central charge against ASC, that a coordinate transformation is "vacuous" and cannot solve a light travel time problem,⁷ applies without modification to his own construction, whose defining move is to relabel η as τ and redraw the initial surface.  The one difference he could point to is that he excises part of the manifold, which is an ontological claim rather than a relabelling.  He never makes that argument.  Nobody has made it for him, so this article will:  it is the best defence of Dennis against Dennis, and it is examined below.

Four models, four authors, one result

This is the finding the article exists to state, and every line of the evidence for it was written by a young-earth author about his own work.

Five entries, four of them written by the model's own author and the fifth by a general relativist checking Humphreys' work.  They agree.  Each proposal differs from standard cosmology in how spacetime is labelled, or in where its initial surface is drawn, and in nothing a telescope can see.

Two things follow, and both matter.

None of these models has been refuted by an observation, and none ever will be.  A critic who goes looking for the observation that kills ASC or kills Dennis's spacetime is wasting his time.  Lisle says this in as many words:  it is "ultimately impossible to disprove a coordinate system."¹⁵  So does Dennis:  the convention "cannot alter physical quantities."⁶  So do Tenev and colleagues:  the choice "has no physical significance at all."²

None of these models has ever predicted an observation either, and for exactly the same reason.  This is the half that gets left out.  If a model makes the same predictions as the standard one, then any observation it predicts, the standard model predicts too.  There is no way to run the reasoning in one direction only.

Lisle's series shows the strain.  In part four he writes that "no physical observation can refute ASC because it will always make the same observable predictions as ESC."¹⁶  In part one and again in part eleven he writes that the ASC model is the only young-earth model that "has made successful predictions regarding astronomical observations," naming massive well-developed galaxies at high redshift.¹⁵⁸  Both cannot be true of the same object.

His defence is a real distinction:  the ASC convention is a coordinate choice, while the ASC model is that convention plus a universe supernaturally created about 6,000 years ago in a mature state.  Fair enough.  But the distinction concedes the point.  Every prediction he scores follows from "created recently, created mature," and not one of them follows from the synchrony convention.  Humphreys' cosmology predicts mature high-redshift galaxies.  So does Hartnett's.  So does Dennis's, since a galaxy created on his hyperboloid is created in whatever state FLRW says it had at that cosmic time.  So does bare created maturity with no convention at all.  The convention is not doing the predicting.

Tenev and colleagues reached the same place from the other side when they wrote that "it is the initial conditions that make the solution possible and not the convention."²  Two of the three parties to this literature have now said it.

What a coordinate change cannot do

Proper time is the same number in every coordinate system

This is the limit, and it is not a matter of opinion.

Proper time is what a clock carried along with an object records.  It is computed by integrating the spacetime interval along the object's own path, and the interval is an invariant.  Relabel the axes however you like;  the integral does not change.  A star's fusion clock, a white dwarf's cooling curve, a radioactive nucleus, an expanding shell of gas:  all of them run on proper time, and all of them are untouched by anything either of these models does.

So neither model removes any evidence of a long history.  Both convert the distant starlight problem into the appearance-of-age problem.  That is not a criticism from outside;  it is what the models are.  Dennis's creation surface cuts a galaxy out of FLRW at whatever cosmic epoch the surface crosses it, so the galaxy begins existence with the state standard cosmology assigns to that epoch:  its stellar populations, its metals, its chemistry, its structure.  Tenev and colleagues place the star's creation event on their hyperboloid, and every subsequent photon reports the star's real history from that moment on, which means the very first photon reported a fully formed star.

The sharpest case: the ring around SN 1987A

The best single test is not a distant galaxy.  It is a supernova in a satellite galaxy of the Milky Way, 168,000 light years away, watched continuously for thirty-nine years.  The SN 1987A article works the whole case;  here is the part these two models cannot absorb.

The explosion lit a ring of gas that the star had shed long before it died.  The ring is measured, not modelled:

"SN 1987A has three circumstellar rings that are part of a complex system of gas that appears to have been ejected by the progenitor some 20,000 years before Sk -69 202 exploded … The inner, equatorial ring is expanding at 10 km s−1;  its true shape is a nearly circular ring of radius 0.67 ly, inclined at 43 degrees to our line of sight."¹¹¹¹

Do the arithmetic.  0.67 light years divided by 10 kilometres per second is about 20,000 years.  Both numbers are measurements:  the radius from the light echo timing, the expansion speed from Doppler shifts in the ring's own emission lines.  No cosmology enters.

Now apply the models.  Tenev and colleagues date the explosion at 5,987 years after creation.²  Dennis's model dates it a few thousand years after creation too, since the creation surface passes just below the star's creation event.  Either way, the star has existed for something under six thousand years.  The ring left the star 20,000 years before the explosion.  Twenty thousand is larger than six thousand.

There is no coordinate transformation that fixes this, because the ring's age is a ratio of two measured quantities on the same object, and both are invariant under relabelling.  The ring must have been created already in flight, 0.67 light years out and moving.  That is created appearance of age, and it is the answer that Danny Faulkner, writing as a young-earth astronomer, called "deceptive" when it was offered for created light.¹⁸¹⁸

This is not an isolated case.  Faulkner's own 2017 table lists seventeen Galactic supernova remnants with ages above 7,000 years, five above a million.¹³  Globular cluster white dwarfs in 47 Tucanae, well inside the Milky Way, give a cooling age of 9.9 ± 0.7 billion years.¹⁹¹⁹  Every one of those clocks runs on proper time.

Where that leaves the two models

Honestly stated:  Dennis and Tenev both succeed at what they set out to do, and what they set out to do is smaller than the problem.

They show that the transit of light is not an obstacle to a recent creation, given the right initial surface.  They do not show, and do not claim to show, that the objects the light came from are young in their own history.  For that, every one of them needs the same thing the oldest and least sophisticated young-earth answer needed:  a universe created with a past it never lived.

The honest case for these two models

It is worth saying plainly what these papers get right, because most of the young-earth literature on distant starlight does not.

They do not require new physics.  No decaying speed of light, no fifth dimension, no modification of the Einstein field equations.  Dennis's spacetime is an exact solution because it is a piece of an exact solution.  Tenev's construction is ordinary special relativity plus a stipulation.  Compare that to Barry Setterfield's decaying c, which fails on the weighted refit of its own data, or John Hartnett's Carmelian cosmology, which its author retracted three times.  The parent article grades all eight.

They state their own limits.  Dennis writes that his models "are still first approximations" and hopes "that young earth physicists trained in GR can take the framework presented and produce models with higher fidelity that fit observational data."¹  Tenev and colleagues write that their solution "does not constitute a complete cosmology" and "does not attempt to explain how creation itself might have happened."²  Neither paper oversells.

They correct their own side.  Tenev and colleagues found and published an arithmetic error that weakened their own argument.²  They also produced a quantitative objection to Humphreys' and Hartnett's time-dilation cosmologies, the blueshift argument, which is the best thing in their evidence section.²

They are mathematically checkable.  Anyone with a graduate course in general relativity can verify Dennis's construction, and anyone who can read a Minkowski diagram can verify Tenev's.  That is rare in this literature and it is worth something.

So a young-earth reader who wants a model that will not collapse under questioning should read these two papers rather than the popular treatments.  They will not be embarrassed by them.

And then they should notice the price.  Both models are unfalsifiable by construction, and their authors say so.  A model that no observation can touch is not a scientific theory in the ordinary sense;  it is a demonstration of consistency.  Dennis's own conclusion, "there can be no unprejudiced objection to this solution,"¹ is meant as a boast and reads as a warning.  A proposal that no evidence can count against is also a proposal that no evidence can count for.

Objections and responses

"This is a change of variables.  You cannot solve a physics problem by renaming things."

Correct, and both sets of authors agree with you.  Dennis says the convention "cannot alter physical quantities."⁶  Lisle says a coordinate change "cannot possibly affect any observational result."¹⁵  Tenev and colleagues say it "has no physical significance at all."²  So nobody in this dispute thinks the relabelling does physical work.

The argument is about what kind of problem distant starlight is.  If it is "how did these photons physically cross ten billion light years in six thousand years," a coordinate cannot help, and Dennis is right.  If it is "when the Hebrew text says the stars were made on the fourth day, what timing rule is it using," then a coordinate is the right shape of answer, and Lisle is right.  That second question is genuine.  The Bible does not state a synchrony convention, and imposing a 1905 convention on a text written millennia earlier is not obviously correct.  Nor is the reverse.  No Hebraist has published on the question, which is the largest gap in this literature.

"Relativity says light travels at c.  An infinite inbound speed is ruled out."

It is not.  Relativity fixes the round-trip speed.  Every measurement ever made of the speed of light has been a round trip, and the one-way split is fixed by definition, as Einstein said.¹²

That said, there is a version of the claim that is ruled out, and a young-earth reader should know the difference.  Andrew Repp made the point in the Creation Research Society Quarterly:  a physically infinite c in one direction "would require a zero ε₀μ₀ in that direction, making either Coulomb's Law or Ampere's Law wildly anisotropic as well.  A mere coordinate redefinition … would not suffer from this problem."²⁰²⁰  Popular presentations sometimes slide from the coordinate claim to the physical one.  The physical one is dead.  Dennis's, Lisle's and Tenev's published versions are all the coordinate claim.

"Dennis proved that Lisle's model contradicts itself."

He did not.  His derivation writes the spacetime interval in the form it takes only when the Einstein convention has already been adopted, then derives the Einstein convention.  Lisle identifies the circle correctly.¹⁴  What survives is Dennis's conclusion, which Lisle accepts anyway.

"Dennis's model is not a relabelling.  He deletes part of spacetime.  That is a real physical difference."

This is the strongest defence of Dennis against his own critique of Lisle, it is correct as far as it goes, and Dennis has never made it.  So here it is, with its limits.

Cutting a region out of a manifold is not a coordinate change.  A spacetime with a boundary and one without are different objects, and no relabelling turns one into the other.  Dennis's construction really does assert something ontological:  the FLRW region below his creation surface never existed.

The trouble is that the excised region is precisely the part nobody can see.  Everything observable lies on Earth's present past light cone.  Dennis's creation surface is drawn to hug that cone from below without meeting it, so the cone lies entirely in the retained piece and the deleted region is unobservable in principle.  The excision is a real difference with no observational consequence, which is a different thing from a coordinate change and lands in the same place.  Dennis says so:  every prediction, including redshifts, is identical.¹

There is one more honest note here.  Cutting spacetime leaves past-directed paths running off the edge after finite elapsed time.  That is exactly the mathematical signature of the big bang singularity itself.  Dennis has not removed the incompleteness from cosmology;  he has moved it, from a single moment at cosmic time zero to a curved surface that reaches back further the further out you look.  Whether a created boundary is more satisfactory than a singular one is a theological judgement, not a physical one, and Dennis presents it as such.

"The missing old supernova remnants prove the galaxy is young."

Not according to Answers in Genesis' own astronomer.  Faulkner reexamined the data in 2017 and reported that "there indeed are old supernova remnants, and that stage three supernova remnants have been identified.  Therefore, this may not be a good argument for recent origin, and I discourage recent creationists from using it."¹³  Tenev and colleagues cite him and answer that the apparently old remnants may reflect "yet unknown processes."²  That answer cannot be confined to one side of the comparison:  the expected counts they use are computed from the same expansion models whose reliability they have just questioned.

"Unfalsifiable means pseudoscientific."

Too fast.  A demonstration that two things are consistent is a legitimate result, and it is what Dennis says he is producing.  Showing that general relativity permits a young-earth reading is worth knowing, and it disposes of the claim that relativity by itself refutes Genesis.

But a consistency proof is not a cosmology, and it should not be sold as one.  The claim these papers can support is "you cannot rule this out."  The claim they cannot support is "the evidence favours this."  Watch for the second claim being made on the strength of the first.

"If these models predict exactly what the big bang predicts, why not accept the big bang?"

That is the right question, and it deserves a straight answer rather than a deflection.

If two models are observationally identical, no measurement can choose between them, and the choice is made on other grounds.  The young-earth advocate is choosing on the grounds of a particular reading of Genesis.  He should say so, and Dennis does:  his model "only differs from conventional solutions in that it uses an initial condition … that supports a young earth."¹  Tenev and colleagues do too.  Their opening sentence under "Proposed Solution" is:  "We infer a solution for the Distant Starlight problem from a straightforward interpretation of Genesis 1 in the context of Special Relativity (Einstein 1905)."²

What follows is that anyone who came to a young earth by way of evidence, rather than by way of the text, is not being served by these models.  They deliver no evidence.  Their whole content is that the text and the evidence can be held together at a cost, and the cost is that the universe must look exactly as old as it looks.

"Earth sitting at a privileged point in spacetime is absurd."

It is unusual, and Tenev and colleagues do not hide it.  They reject the cosmological principle explicitly, calling it "an atheistic assumption rooted in the belief that the cosmos is not designed."²  That characterisation is too strong.  The cosmological principle began as a simplifying assumption and has since been tested:  the microwave background is isotropic to about one part in one hundred thousand once the dipole is removed, and large surveys find the galaxy distribution homogeneous above about 100 megaparsecs.  It is not a bare presupposition any more.  The microwave background article and the large-scale structure article give those measurements with their uncertainties.

The honest form of their position is not that the principle is unmotivated, but that the observations constrain isotropy about us far better than they constrain homogeneity, and a strictly Earth-centred model can survive.  The article on cosmic size and young-earth cosmologies takes that up.  It is a live position and it is expensive.

"Are you just repeating what secular critics say about creationists?"

Check the citations.  The finding that these models predict nothing new comes from the models' own authors, in their own papers.¹²  The retirement of the supernova-remnant argument is by the staff astronomer at Answers in Genesis in an Answers in Genesis journal.¹³  The warning that popular presentations state a version of the light-speed claim that electromagnetism rules out is in the Creation Research Society Quarterly.²⁰  The judgement that created appearance of age "is deceptive" is Faulkner's.¹⁸  The charge that Lisle's exegetical justification is "a rather weak justification" is Tenev, Baumgardner and Horstemeyer's.²  Every load-bearing criticism in this article was published by a young-earth creationist.

Sources

The two primary papers are the 2018 International Conference on Creationism proceedings volume, both open access through Cedarville University's repository.  Dennis's is pages 14–35;  Tenev, Baumgardner and Horstemeyer's is pages 82–94.  Both were read in full from the publisher's PDFs.  The 2024 exchange is in Answers Research Journal volume 17, also open access.  Lisle's replies are self-published on the Biblical Science Institute site and were harvested from the live pages.

Every quotation below was taken from the source bytes.  The two conference papers are two-column, and page numbers were established by extracting each printed page separately and reading the running head, because a plain extraction interleaves the columns.