GAE hypothesis: some student confusion

This post is primarily for @swamidass but I thought perhaps others might be interested.

I have a general education undergraduate class going through GAE this semester. I’m using the book as an adjacent “science and faith” component to the course, not as the main content. About 2/3 of the students are Christians who predominately assume YEC. The other 1/3 are non-religious students (atheists/agnostics/“Nones”).

I had the students read through the genealogical hypothesis in Chapter 2 (pg. 25):

  1. They [Adam and Eve] lived recently in the Middle East.
  2. They are genealogical ancestors of everyone.
  3. They are de novo created.
  4. Interbreeding between their lineage and other.
  5. No additional miracles allowed.
  6. The two findings of evolutionary science.

I asked the students to indicate if they found any of the six elements above, and while many said they found it easy to understand, I noticed two trends:

  1. The students generally had no idea what de novo meant. Many of them connected it with ancestry, i.e. that Adam and Eve were created by God and the first humans. They took de novo to mean “beginning of humanity” rather than “created out of nothing” that I normally think of.
  2. Several of the students commented that they didn’t understand 5 (no additional miracles). They wondered why @swamidass would put that in the GAE hypothesis and didn’t seem to understand it’s purpose. I think this reflected them having a hard time understanding what a thought experiment is and why methodological naturalism is important.

So I wonder if you all have any good suggestions on how to address these two confusions that is appropriate for non-science students.


Sounds like a fun class!

For this, emphasize the text of Scripture (“from the dust” and “from a rib”) and the phrase “without parents.”

The key point there is that a miracle is allowed in the hypothesis, but not in the assessment of the hypothesis (see here: Does the GAE violate methodological naturalism?).

You want them to understand that rule #5 is required for assessment of the evidence, otherwise we can just start invoking miracles willy nilly to explain away difficult data. That would pollute our analysis and prevent us from being able to rule out any hypothesis, so it just is not allowed in scientific analysis. If we can posit ad hoc miracles, there is just no way to test a hypothesis in a rational way.

One pedagogic strategy might be to consider Coyne’s quote in the first chapter. He considers the Virgin Birth, the Resurrection, and Adam and Eve. How does he does he make the (correct) scientific claim that evidence doesn’t rule out the Resurrection and the Virgin Birth?

Another pedagogic strategy would be to consider an argument that posits ad hoc miracles to explain data away. You could specify a hypothesis that is falsified by evidence (such that the students agree this to be the case), and then ask them to posit miracles that give an alternate explanation. Alternatively, you could ask them to falsify a clearly false statement (e.g. that the class is now on the Moon), and posit miracles to dismiss all the evidence they present against this statement.

The point is that we can’t make evidential assessments of a hypothesis if we are allowing miracles to do any explanatory work for us. If we did, we would not be doing a fair assessment.



I another distinction is between “potentially attested” miracles, e.g. which are arguably stated in Scripture, versus “unattested” miracles that only arise as a way of explaining away difficult evidence. Those potentially attested miracles are interesting to place in an “improper” hypothesis, but the unattested miracles are less interesting there is no potential warrant for them.

This shows why unattested claims, such as a teacup orbiting the planet or unicorns on the other side of the moon. There is no evidence against these hypothesis, but there also is no potential warrant for them.


I am also using GAE in my Human Biology class. I avoid the phrase “thought experiment.”. Instead, I say “Let’s pretend this happened. What would that look like?”. I get them to make predictions, then follow up with “how could we test that?”. It’s the same idea, but without the terminology.
My students have been struggling with a different idea - ghost ancestors. I had to explain that not only does each parent only contribute half of the DNA, but a lot of the DNA is identical, meaning we can’t tell who it came from. I hope I got that right. Sometimes I have to simplify ideas so my students can understand them.
This leads me to saying things that aren’t 100% technically correct but convey the main point.


@jordan (@swamidass )

Should Point [5] be re-worded? Naturally, we have to allow for the one-off miracles of the miraculous birth of Jesus and the miraculous resurrection of Jesus.

And more concisely, any one-off miracles that don’t over-turn all of science are generally permissable!

@nlents do you have any suggestions here?

@swamidass, @Jordan

This point makes sense … ASSUMING that there is distinction between natural processes that are providential and seemingly super-natural processes that would be categorized as miracles.

If one is a Christian, there is no reason to embargo discussions of providential-but-natural events.


Don’t forget that with each generation, the number of potential ancestors increases geometrically, but there are still only 46 chromosomes to express the genetic “baggage” of these ancestors. Even with robust recombination (where branches of chromosomes are exchanged in unexpected ways), the odds of having each generation of ancestor’s reflected in the 46 chromosomes decreases dramatically!

I love this image and I think it helps. This is result of a genetic analysis of a single sperm cell, showing in alternating colors the “maternal” and “paternal” DNA that this sperm would pass on to the next generation, if it were to fuse with an egg. Notice that there are “chunks” of DNA that get passed on, at random, meaning lots of chunks left out. Do this over 4-5 generations and you end up with some ancestors passing on nothing in any chunk.

Sorry, this was quick. But does it help?


I think this is where there is some confusion. @swamidass is saying that for purposes of the thought experiment and clarity of analysis, the only miracle considered in GAE is going to be the de novo creation of Adam & Eve in the first place. From a “scientific” standpoint this makes sense since if your hypothesis includes an arbitrary number of events that are not assessable via scientific evidence, then the whole hypothesis becomes pretty pointless.

Take, for instance, the Tasmania critique. If the hypothesis allows for a miracle at each “difficult” spot then the hypothesis could literally explain any data, and if a hypothesis can explain any data it can’t distinguish between anything. So I think point 5 (no more miracles allowed) is really important for GAE as a thought experiment to have any sort of scientific credibility, but that doesn’t mean that Josh, or anyone, is saying that there couldn’t have been additional miracles.

I notice that there are 22 chromosomes here, presumably leaving out the sex chromosome, and only 24 crossover sites. Most chromosomes have one crossover, Chromosome 2 has 3, Chromosome 10 two, and Chromosome 16 has none. This seems to me like a low count. But I don’t know that much about human recombination.


When did the “Tasmania critique” become so bizarre?

You are calling a natural process a miracle because it was used by God. For purposes of this discussion, this is erroneous. And if Joshua used the terms in this way, he was in error as well.

For Christians, providential events are all over the place… without imagining any super-natural activity.

You don’t have to shout. And you are claiming the right to define what Christians think, while depriving @Jordan and @swamidass of that right. You have never been able to clearly define what providence is and how it differs from miracles on one hand and from natural events on the other. Best to keep GAE free from all that jargon.

1 Like

Exactly. This is a methodological rule, not a metaphysical premise.

Exactly. We are not presuming a closed universe free of miracles, but we are refraining from using miracles to explain away difficult data.

This is important because @gbrooks9 often appeals to miracles on Tasmania, but I explicitly forbid this in my analysis. This does not rule out the possibility of miracles, but I do not rely on them in my analysis and I would actually prefer if @gbrooks9 did not make the appeal.

1 Like


Haven’t I been rather CONSISTENT about referring to these Tasmanian events as providential, rather than miraculous? Why are you referring to all “providential” events as “miraculous”?

How does this category of “providential events”, built entirely of natural processes, become such a chunk of kryptonite to you, Joshua?

@gbrooks9, yes but you started with:

I’m saying that it’s very reasonable to not allow what people would understand as a miracle within the GAE hypothesis in order to make the argument clear and scientifically defensible. That doesn’t mean that miracles aren’t allowed.

Now, Providence is something entirely different and not something that many people will see the nuance of. I still think that appealing to Providence will only muddy the waters.

1 Like

Exactly. As I stated before:

I wonder if @AndyWalsh could help the students with some web-based simulations, or if any one knows where one could be found.

This is from actual data. In humans, all chromosomes typically have at least one (the ones that appear to have none could either have two that were very close together, escaping detection, or it happened very near the extreme ends where no markers are found) because zero crossovers leads to meiotic errors. In humans, two is somewhat rare, and three is very rare. The sex chromosomes are not shown because they don’t cross over (in males, and this is a sperm cell).


I’d rather say that in allowing for miracles for each “difficult” spot renders any statements of evidential status entirely meaningless.

Take scientific YEC, for example. They claim two contradictory things:

  1. The evidence demonstrates their position.
  2. The evidence must be interpreted assuming their position is correct (often invoking miracles)

If the evidence demonstrated their position, all we would need is a plain reading of the evidence, and we we would not need interpret the evidence assuming they were correct and invoking miracles as needed. Rather, they would be better off saying that,

  1. The evidence is not necessarily in conflict with their position if we allow for God’s action, because science does not consider God’s action.
  2. A plain reading of the evidence would seem to point a different direction, which explains why most scientists disagree with them.

That, it seems, is a much more coherent position.