Mark's use of Matthew and Luke

[ProfessorDavey]

So if some were accurate at each stage of the transmission, then the original text, or most of it, would have been transmitted to us.

Not true.

Try this simple example. Imagine a document which needs to be copied every 20 years as the original falls apart (as is the case usually with papyrus). Lets say the likelihood of the copying being completely accurate is 50% - so 50% of the tme the copy is identical to the original, 50% of the time errors creep in. Now I think this is being massively optimistic in the context of copying the gospels in the early decades as this was done by people with no real skills or training. But regardless lets work with the 50% assumption.

So fast forward 200 years (the earliest when we start to see any meaningful extant gospel copies) - so 10 generations of copy (once every 20 years). The likelihood that the 10th generation copy is identical to the original is less than one in a thousand.

And this is a simply, linear example. The early transmission of the gospels is much more complex than that - with multiple copies created and distributed across a wide geographic region, copied more and spread more. So by the time you've reached 200 years after the original you will have hundreds, maybe thousands of copies. If we had all of those copies from 250AD onwards then perhaps one of them would be the same as the original (quite likely none of them will be) but how would you know which is the same and which are different - you couldn't, unless you have all the prior copies back to the original (which we don't). But of course we do not have all the copies - we have just a tiny proportion of them so the chances that just by luck the copies that happed to survive happen also to be the incredibly rare ones that are faithful to the original is, frankly, vanishingly small.

Now I fully accept that all I am talking about here is changes - not the significance of those changes. In that context I think we can be pretty confident that none of the extant fragments from decades and centuries after the gospels were written are vanishingly unlikely to be identical to the original. But of course changes can be minor and irrelevant to meaning or major edits, omissions and deletions which have a marked change in meaning and interpretation. We can be pretty confident that there are changes and given that the extant copies we have from the 4thC have major differences - whole sections in some that aren't in others, significant changes in wording then we can conclude that at least some of the changes from the original are pretty major.

[jeremyp]

So if some were accurate at each stage of the transmission, then the original text, or most of it, would have been transmitted to us.
I think Jeremy said earlier that having three gospels which most of the time agree on the details, is itself evidence that they have been transmitted accurately.

Fairly accurately.

If there had been major errors then they wouldn't agree.

If I gave you the impression I meant that, I apologise.

What I was trying to refute was the assertion that we do not know what they said. PD implies that we can have no idea at all what was in the original gospels. However, that would require the ability of copyists to make accurate copies over the two or three hundred years for which we have no complete manuscripts to be much worse than for the succeeding thousand years.

The Codex Sinaiticus was written in the fourth century and has a number of variations over what we would think of as being in the Bible and yet it is recognisable as being the same Bible as existed a thousand years later but with some variations. I think it is a reasonable assumption that the original documents (for the NT at least) were very similar to the Codex Sinaiticus.

Furthermore, the Synoptic gospels probably diverged and became separate documents very early in their history- before the end of the first century. After that, they would have evolved independently, but we can still identify the sections in Matthew and Luke that were copied from Mark (or vice versa) and we can still identify the sections where Matthew and Luke had a common source.

All of those factors together tell us that we do have a pretty good idea of what was in the original gospels, contrary to what PD asserted.

And even if some parts were added by later editors, often these parts agree with another gospel: for example, Mark 16:9-20 agree with another gospel or Acts. Agreed?

The trouble with that specific example is that whoever wrote the ending of Mark had access to the other gospels and Acts. It's pretty clear that it is written as a précis of the post resurrection stories in those books. i.e. it is not independent.

[Spud]

Not true.

Try this simple example. Imagine a document which needs to be copied every 20 years as the original falls apart (as is the case usually with papyrus). Lets say the likelihood of the copying being completely accurate is 50% - so 50% of the tme the copy is identical to the original, 50% of the time errors creep in. Now I think this is being massively optimistic in the context of copying the gospels in the early decades as this was done by people with no real skills or training. But regardless lets work with the 50% assumption.

So fast forward 200 years (the earliest when we start to see any meaningful extant gospel copies) - so 10 generations of copy (once every 20 years). The likelihood that the 10th generation copy is identical to the original is less than one in a thousand.

Yep, good maths, not sure I agree with the theory. Each time something is copied, different mistakes will be made. Suppose Mark gave his document to a church and moved on. Two people made copies: one missed out one word, the other copied the same word correctly but missed out another. The original perished so that they only had the two copies, and they found that there were two differences in the wording between them. Unless the context for each difference made it clear when a word had been missed out (which is likely), they wouldn't know which copy was accurate in the two places where they were different. But if three copies had been made then it's probable that two would have the original wording and one a different wording, hence they would know that the rendering given by the two would most likely be correct. If ten copies were made of the original, the ability to distinguish where errors were creeping in would be a lot better when it came to making the next generation of copies.

[jeremyp]

Yep, good maths, not sure I agree with the theory. Each time something is copied, different mistakes will be made. Suppose Mark gave his document to a church and moved on. Two people made copies: one missed out one word, the other copied the same word correctly but missed out another. The original perished so that they only had the two copies, and they found that there were two differences in the wording between them. Unless the context for each difference made it clear when a word had been missed out (which is likely), they wouldn't know which copy was accurate in the two places where they were different. But if three copies had been made then it's probable that two would have the original wording and one a different wording, hence they would know that the rendering given by the two would most likely be correct. If ten copies were made of the original, the ability to distinguish where errors were creeping in would be a lot better when it came to making the next generation of copies.

The flaw in your argument is that the ten copies would almost certainly be distributed around the Roman Empire. The person who makes a second generation from one of the ten copies won't have the other nine to check against and probably won't have the original either because they's probably all be several days or weeks journey away.

[ProfessorDavey]

I think Jeremy said earlier that having three gospels which most of the time agree on the details, is itself evidence that they have been transmitted accurately. If there had been major errors then they wouldn't agree. And even if some parts were added by later editors, often these parts agree with another gospel: for example, Mark 16:9-20 agree with another gospel or Acts. Agreed?

That might be an argument if the gospels were being transmitted independently of each other and with no opportunity for one to be aligned with the other during the transmission phase.

But that isn't the case - from pretty early the gospels (and other texts) were being collated into a folio - effectively what we now know as the New Testament. One of the most important early papyrus (p45), likely from earlier than 300AD contains fragments of all the gospels and acts. Now there is very little text to go on - it is estimated that there would have originally been 200 pages and 170 of those are completely absent, with the rest so damaged that I don't think an entire single line is intact. So if cannot tell us much about the text as over 90% is missing, but it does tell us that transmission of the gospels even at this early stage wasn't independent, but in a collected folio.

This means that the copyists won't have just been copying Mark or Matthew, but all four gospels. This means that it would have been very easy (and perhaps even directed) to gently manipulate the gospel texts to create sufficient alignment to provide coherence gospel to gospel. Whether they originally were so aligned is anyone's guess as we don't have the earliest texts (the autograph) - that a couple of hundred years later there is alignment doesn't really help us as the gospels were being transmitted and copied together.

[ProfessorDavey]

The flaw in your argument is that the ten copies would almost certainly be distributed around the Roman Empire. The person who makes a second generation from one of the ten copies won't have the other nine to check against and probably won't have the original either because they's probably all be several days or weeks journey away.

I'm not sure if anyone has estimated the numbers of copies circulating between the time of the original autographs (perhaps 70-90AD) and the point when we finally get complete, or near complete versions available to us - around 350AD.

Given that around 70 'early' (before around 400AD) papyri have survived and papyrus very rarely lasts beyond a few decades, I think these 70 are probably just the tip of the iceberg.

[Spud]

That might be an argument if the gospels were being transmitted independently of each other and with no opportunity for one to be aligned with the other during the transmission phase.

But that isn't the case - from pretty early the gospels (and other texts) were being collated into a folio - effectively what we now know as the New Testament. One of the most important early papyrus (p45), likely from earlier than 300AD contains fragments of all the gospels and acts. Now there is very little text to go on - it is estimated that there would have originally been 200 pages and 170 of those are completely absent, with the rest so damaged that I don't think an entire single line is intact. So if cannot tell us much about the text as over 90% is missing, but it does tell us that transmission of the gospels even at this early stage wasn't independent, but in a collected folio.

This means that the copyists won't have just been copying Mark or Matthew, but all four gospels. This means that it would have been very easy (and perhaps even directed) to gently manipulate the gospel texts to create sufficient alignment to provide coherence gospel to gospel. Whether they originally were so aligned is anyone's guess as we don't have the earliest texts (the autograph) - that a couple of hundred years later there is alignment doesn't really help us as the gospels were being transmitted and copied together.

But there are still conradicting details in the gospels. Why weren't these harmonized?

[ProfessorDavey]

But there are still conradicting details in the gospels. Why weren't these harmonized?

Not everything is going to be changed or you'd end up with one, rather than four, gospels - having four gospels would have helped the early church communicate their message as the notion of four narratives inherently implies independent corroboration (when of course three of those gospels, at least, aren't independent of each other. Also the synoptic gospels are considered to have been written for different audiences, so useful to keep some differences in tone even if the main elements become harmonised.

I think also you get to a point when an orthodox version of the NT is established - but that took some time to occur.

But we have clear evidence of the harmonisation - so the early versions of Mark have no resurrection witness narratives, yet Matthew and Luke of the same time do. Fast forward a little and Mark has been harmonised with the other synoptic gospels by the addition of extra verses to chapter 16 to include post-resurrection appearances.

The fundamental point is that by the time we have any meaningful extant textual evidence of the gospels they are not independent (if they ever were - that we don't know) - they are being copied and transmitted alongside each other.

[Spud]

Not everything is going to be changed or you'd end up with one, rather than four, gospels - having four gospels would have helped the early church communicate their message as the notion of four narratives inherently implies independent corroboration (when of course three of those gospels, at least, aren't independent of each other.

Yes, if everything was changed there would only be one gospel. I find it hard to believe that details such as whether Jairus' daughter was alive when her Dad first came to Jesus would have been deliberately manipulated. It is possible that the contrast between Matthew's and Mark/Luke's versions could be due to copying error, but even then the essence of the story doesn't change - Jesus brought her back to life. Similarly with most other contradictions, they don't change the essence of the story.

Also the synoptic gospels are considered to have been written for different audiences, so useful to keep some differences in tone even if the main elements become harmonised.

Fair enough.

I think also you get to a point when an orthodox version of the NT is established - but that took some time to occur.

From what I can tell, not very long.

But we have clear evidence of the harmonisation - so the early versions of Mark have no resurrection witness narratives, yet Matthew and Luke of the same time do. Fast forward a little and Mark has been harmonised with the other synoptic gospels by the addition of extra verses to chapter 16 to include post-resurrection appearances.

That is evidence of someone, possibly Mark, finishing off the story at a later date, it doesn't mean Mark wasn't aware of any resurrection appearances. The NT is clear that he was an active member of the early believers.
That's one example, there are others where some verses or words are omitted by the early manuscripts, but they don't change the meaning of the text.

The fundamental point is that by the time we have any meaningful extant textual evidence of the gospels they are not independent (if they ever were - that we don't know) - they are being copied and transmitted alongside each other.

The retired homicide detective Warner Wallace answers this by citing the writings of the early church fathers who taught the same message as that of the earliest of our manuscripts. He says there's a chain of students including Polycarp, Irenaeus etc linking back to the apostles, and that their message did not change over time, which it would have if the original message was different.

[ProfessorDavey]

That is evidence of someone, possibly Mark, finishing off the story at a later date ...

Blimey he must have lived to a ripe old age, noting that the original is considered to have been written in about 70AD and all copies up to about 400AD don't include the additional verses, so they were likely added in the 5thC. So if Mark added them he'd have been about 400 years old by then!

And of course it goes without saying that we have no idea who the author (or authors) of the gospel attributed to Mark was (or were).

[Spud]

Blimey he must have lived to a ripe old age, noting that the original is considered to have been written in about 70AD and all copies up to about 400AD don't include the additional verses, so they were likely added in the 5thC. So if Mark added them he'd have been about 400 years old by then!

And of course it goes without saying that we have no idea who the author (or authors) of the gospel attributed to Mark was (or were).

I have just read that codex Vaticanus was aware of the longer ending. It apparently has a blank space at the end, the only such space it contains.
Irenaeus quoted Mark 16:19 around 200 AD:

Also, towards the conclusion of his Gospel, Mark says: "So then, after the Lord Jesus had spoken to them, He was received up into heaven, and sitteth on the right hand of God;

http://gnosis.org/library/advh3.htm
Note in this writing, all the other quotes from the gospels which agree with our text.

[jeremyp]

I have just read that codex Vaticanus was aware of the longer ending. It apparently has a blank space at the end, the only such space it contains.
Irenaeus quoted Mark 16:19 around 200 AD: http://gnosis.org/library/advh3.htm
Note in this writing, all the other quotes from the gospels which agree with our text.

Maybe Mark 16:19 quotes from Irenaeus.

[Spud]

You need to understand that neither phrase is the original (the autograph), nor the archetype but an English translation of an earlier version which will have been written itself centuries after the original first appeared.

Better, "it is not certain that either phrase is the original".

One reason why we could be looking at the original phrases is that this kind of discrepancy, where Matthew makes more sense than Mark and it's hard to see why Mark would have written as he did when writing before Matthew, occurs often. Here is another:

And having stripped Him, they put a scarlet robe around Him.
And having twisted together a crown of thorns, they put it on His head, and a reed in His right hand - Matthew 27:28,29

And they put on Him purple, and having twisted together a crown of thorns, they placed it on (literally, 'around') Him - Mark 15:17

Matthew uses the verb peritithemi (to place around) in a coherent way, Mark's use is less coherent: he says that they placed the crown of thorns around him, substituting 'him' for 'his head', thinking it meant 'on him'. Mark also uses the word 'reed' less coherently, mentioning it only when the soldiers strike Jesus with it.

It is more likely that Mark used 'place around' because he had read it in Matthew's text, than that, writing before Matthew, it came into Mark's mind some other way.

It's hard not to see a pattern of error in Mark's work that must have come about through a sort of editorial fatigue on his part.

[Nearly Sane]

Better, "it is not certain that either phrase is the original".

Except it is entirely sure since the original was not written in English.

[jeremyp]

It's hard not to see a pattern of error in Mark's work that must have come about through a sort of editorial fatigue on his part.

That's not how editorial fatigue works. Editorial fatigue occurs when you make a deliberate change to a text but you do not carry it consistently through the whole text.

[Spud]

Fairly accurately.
If I gave you the impression I meant that, I apologise.

What I was trying to refute was the assertion that we do not know what they said. PD implies that we can have no idea at all what was in the original gospels. However, that would require the ability of copyists to make accurate copies over the two or three hundred years for which we have no complete manuscripts to be much worse than for the succeeding thousand years.

The Codex Sinaiticus was written in the fourth century and has a number of variations over what we would think of as being in the Bible and yet it is recognisable as being the same Bible as existed a thousand years later but with some variations. I think it is a reasonable assumption that the original documents (for the NT at least) were very similar to the Codex Sinaiticus.

Furthermore, the Synoptic gospels probably diverged and became separate documents very early in their history- before the end of the first century. After that, they would have evolved independently, but we can still identify the sections in Matthew and Luke that were copied from Mark (or vice versa) and we can still identify the sections where Matthew and Luke had a common source.

All of those factors together tell us that we do have a pretty good idea of what was in the original gospels, contrary to what PD asserted.

I forgot to thank you for this.

The trouble with that specific example is that whoever wrote the ending of Mark had access to the other gospels and Acts. It's pretty clear that it is written as a précis of the post resurrection stories in those books. i.e. it is not independent.

Yes and just to note that some think much of the rest of Mark shows signs of having Matthew and Luke as its sources!

[Spud]

That's not how editorial fatigue works. Editorial fatigue occurs when you make a deliberate change to a text but you do not carry it consistently through the whole text.

OK.

I'm not sure what it would be called, but there is still a pattern of error in Mark, where because he makes deliberate changes, his sentences become less coherent than his source (Matthew or Luke, according to Markan dependence).

PD quoted Bart Ehrman as saying, I think, that when there are two texts describing the same event, the one that is the more difficult reading is the original. I wonder if this applies when one of the texts is written by an eyewitness? If Matthew was written by eyewitnesses then the author would relate events as he recalls them and they would be likely to be coherent. Someone using Matthew as a source who was not an eyewitness could slip into the habit of changing bits and disturbing the flow of thought.

[jeremyp]

PD quoted Bart Ehrman as saying, I think, that when there are two texts describing the same event, the one that is the more difficult reading is the original.

"Difficult reading" in this instance means less in accord with received theology, not hard to read or clumsily written. For example, a reading that shows Jesus to be fallible might be more difficult than one that shows him to be infallible.

Also, the rule is probabilistic. It's not guaranteed that the difficult reading is the easier one, only more likely.

I wonder if this applies when one of the texts is written by an eyewitness? If Matthew was written by eyewitnesses

Matthew wasn't written by an eye witness.

then the author would relate events as he recalls them and they would be likely to be coherent. Someone using Matthew as a source who was not an eyewitness could slip into the habit of changing bits and disturbing the flow of thought.

Not necessarily. The memory might not be coherent and the next person could smooth out the inconsistencies.

[ProfessorDavey]

"Difficult reading" in this instance means less in accord with received theology, not hard to read or clumsily written. For example, a reading that shows Jesus to be fallible might be more difficult than one that shows him to be infallible.

Actually Ehrmann's view on what 'difficult reading' means goes beyond that -in his words:

Therefore, when we have two forms of a text, one that would have been troubling to scribes—for example, one that is possibly contradictory to another passage or grammatically inelegant or theologically problematic—and one that would not have been as troubling, it is the former form of the text, the one that is more “difficult,” that is more likely to be original. That is, since scribes were far more likely to have corrected problems than to have created them, the comparatively smooth, consistent, harmonious, and orthodox readings are more likely to have been created by scribes.

So it also includes clunky grammar, disjointed narratives in addition to elements that may be problematic theologically.

[Spud]

Actually Ehrmann's view on what 'difficult reading' means goes beyond that -in his words:

Therefore, when we have two forms of a text, one that would have been troubling to scribes—for example, one that is possibly contradictory to another passage or grammatically inelegant or theologically problematic—and one that would not have been as troubling, it is the former form of the text, the one that is more “difficult,” that is more likely to be original. That is, since scribes were far more likely to have corrected problems than to have created them, the comparatively smooth, consistent, harmonious, and orthodox readings are more likely to have been created by scribes.

So it also includes clunky grammar, disjointed narratives in addition to elements that may be problematic theologically.

But then when one tries to explain how the inelegant or disjointed narrative came to be like that, the only way I can think of other than if he was borrowing it from his source and rearranging some of it, is that he writes things as they come to him and doesn't self-edit. I would like to see any examples Ehrman can give for his claim.

[ProfessorDavey]

But then when one tries to explain how the inelegant or disjointed narrative came to be like that, the only way I can think of other than if he was borrowing it from his source and rearranging some of it, is that he writes things as they come to him and doesn't self-edit. I would like to see any examples Ehrman can give for his claim.

The most obvious reason being that the author was ramming together bits and pieces from multiple sources which were themselves transmitted to the author in written or oral format in different styles. The narrative equivalent of a cut and shut car. The earlier versions would be disjointed but over time and rewriting the clunkiness gets smoothed over to create a more coherent narrative style.

[Spud]

The most obvious reason being that the author was ramming together bits and pieces from multiple sources which were themselves transmitted to the author in written or oral format in different styles. The narrative equivalent of a cut and shut car. The earlier versions would be disjointed but over time and rewriting the clunkiness gets smoothed over to create a more coherent narrative style.

That sounds something like the Griesbach hypothesis: Mark used two written sources (Matthew and Luke) and added to them information from other oral sources (eg names such as Levi being the son of Alphaeus,  Bartimaeus). Matthew and Luke have material in common but in a different order; Mark always follows the order of one or the other or both.

[ProfessorDavey]

That sounds something like the Griesbach hypothesis: Mark used two written sources (Matthew and Luke) and added to them information from other oral sources (eg names such as Levi being the son of Alphaeus,  Bartimaeus). Matthew and Luke have material in common but in a different order; Mark always follows the order of one or the other or both.

Ehrmann in making his comment is not focussing on who borrowed from who. Rather he is using his 'difficult' criterion as one of a range of tools to help scholars determine which of two (or more) versions of the same document is likely to be the closer to the original.

The problem with the who borrowed from who discussions is that we do not have independent versions of the gospels - pretty well everything we have is from dates when the gospels were being copied together and transmitted together in a collected folio (what we now know as the New Testament) - so throughout copying and transmission generation to generation of versions the copyists will have all of the gospels available to them.

[Spud]

I've been working on the assumption that what we have now is  close to the original...

[ProfessorDavey]

I've been working on the assumption that what we have now is  close to the original...

Beyond wishful thinking and special pleading what justification have you for that assertion.

We have virtually nothing in terms of text for about 150 years and most sections of the gospels aren't available to us until the 4thC. We do not, and cannot know, how the texts we have, even the earliest texts available compared to the 'original'. Indeed many scholars aren't sure the notion of an 'original' is even valid, as there were likely a number of copies produced at the very earliest stage, which may not have been the same, nor may they be faithful to the intentions of the purported author, as they were likely written down by scribes, not be the purported author (or authors) himself (or themselves).

In reality all we can really say is that we know how the texts of the gospels had settled by 4thC (which includes many, many variations) - we can use tools to predict which of those variants is closer to the original compared to others - however that is only relative - variant X is likely closer to the original than variant Y. Whether variant X is identical to, similar to, or massively different from the original cannot be determined even if we think it is closer than variant