The Size and Status of Jesus’ Family

The Size and Status of Jesus’ Family June 29, 2014

The depiction of the size of Jesus’ family in Mark 6:3 puts the number of people in the family above the average family size estimated by historians for his time and place in history.

I had this come up in connection with a blundered attempt by a mythicist (who apparently doesn’t understand what an average is) to use an estimate of the maximum family size in Iron Age I proto-Israelite settlements to try to argue that the information about the size of Jesus’ family could not be correct. Yes, seriously.

But once my thoughts turned to this topic, it led to some interesting and more serious questions and reflections. I’ve argued on other grounds that Jesus’ family appears to have had some significant social status. I wonder whether an above-average sized family is more likely to have been of higher status, with their better economic status and/or social connections having provided the opportunity to see more children live to adulthood. Some books and articles I’ve read suggest that a large family was a sign of status, while others suggest that the wealthy tended to have smaller families. If anyone has relevant archaeological, socioeconomic, and other perspectives to offer on this, please do share them!

"I just read Neil Godfrey's take on this post. He wrote:(1) A stark illustration (not ..."

What Jesus Learned from Mary of ..."
"The posted comic reminds me a little of the argument from achievement: If God simply ..."

Batman and Theology #CFP
"Anyway, as I said I was hoping we could move on from that particular confusion. ..."

Challenge to Fundamentalist Bible Readers
"One player I just found online summed up the battle in this way:On my return ..."

Sci-Fi, Superhero, Game, and Cinematic #CFP ..."

Browse Our Archives

Follow Us!

TRENDING AT PATHEOS Progressive Christian
What Are Your Thoughts?leave a comment
  • MattB

    Hello, Dr. McGrath!

    I found something interesting about family life in biblical times at this link(

    It seems that the mythicist you encountered doesn’t really understand much about family life during 1st century Palestine. It would seem that archaeology would show that much of what historians and scholars know about families during Jesus’ time would have had extended relatives live under the same roof.

  • Anonymous

    Why the “yes seriously”? The economy of Palestine did stay the same for thousands of years, although your article on Jesus’s social standing might be compelling evidence against it.

    I’ll don’t have access to the other article but will let the blogger know, who might be able to read it.

    • Michael Wilson

      Its difficult to use familly size in periods seperated by a millinium as a sure stabdard, especially iron age one, which was a period of great change. I suspect that your average family size was probably close though, your right, things didn’t change that much over the span of antiquity. But averages are not absolutes nor should much faith be placed in maximum estimated sizes for a family. Nuclear famillies of more than 8 are not unheard of in the past any more than modernity. It is certainly on the large side but hardly improbable. And keep in mind that all of them need not have been under the same roof, the older kids may well have been adults when the youngest were born.

      • Anonymous

        We know that the family averages for both eras did stay the same and that economic conditions likely the same as well so it’s very likely that the limit stayed the same well. I’d say “moving out” is balanced by low life expectancy.

        Then again, if the Jesus family had substantial wealth (as I take it McGrath is arguing) then it might offset it completely.

        • Michael Wilson

          The economic conditions in the first century Palestine were a lot different than Iron I, but even with a much larger population, I doubt that the household average would have been much higher. More important is that averages mean sone had less and some had more. Some poor slobs may have had 10 kids while a lot of people nay gave lived by themselves. That Jesus had around six siblings doesn’t really say much concrete about him. The figure isnt so large as to be suspicious nor would it be proof of any wealth. More likely, yes, large for the time, sure, but I would bet that it was not unheard of to have a poor laborer with a large family.

          • arcseconds

            The argument given on the blog, and introduced by Anonymous-of-the-Escher-Self-Portrait to Exploring Our Matrix here is that 7 is the upper limit to family sizes, and therefore Jesus’s family is impossible.

            I don’t find this argument to be particularly compelling myself, mind…

    • Erp

      I seriously doubt the economy of Palestine was the same for thousands of years. Among other things for better or worst in Jesus’s time it was hooked into the Roman empire.

      The Oxford Handbooks Online (not necessarily visible to all) have some pointers but admit we know little and what little we know is untrustworthy. More for Roman Egypt for which some census records still exist. A Handbook article (Children in Roman Egypt, April Pudsey)
      has one table that includes a household with 4 brothers all over the age of 35 (131-Pr-1), note that any surviving sisters would have married and gone to other households. Several of the families listed have 5 children (but again any daughters would like marry out sometime in their teens and some sons might not be present for other reasons). None with 7 but the article wasn’t addressing max family size and didn’t include all data known (and we only have a very small portion of the actual census).

  • Michael Wilson

    A quick search didn’t turn up anything on familly size by economic status any where in the early Roman empire. How ever, I did find a mention in Steven Pinker’s “the better angels of our natures” p.425 regarding births among medival families of avergages of 5.1 for wealthy, 2.9 for middle class and 1.8 for the poor. My guess, and I would have to substantiate this, is that poorer families would be smaller since poor nutrition would reduce the number of births, fertillity, and life expectancy. Of course averages don’t address specific examples and any one familly might have been especially lucky or unlucky.

    • Though I’d never advise to trust Pinker for all his assertions in his book (like socialism being a force against the enlightenment rather than springing OUT OF the enlightenment) I think he should be reliable for that one.

  • Anonymous
    • Michael Wilson

      I wouldn’t doubt that it is impossible for an average family to be lager than 7, as you site, but again that’s average, not high or low. Its unlikely that Jesus’ family had 9+ people but it is also unlikely that Josephus could read and write, only one % of the ancient population could. Its unlikely that Obama is president, only one person in 300 million in America is during any 4 year span. You see my pint? I don’t know how many siblings he had. I think we can all be sure that if he was a person, he had a mother and father, and the statement from Paul that he had a brothers makes sense too. Lots of people probably had brothers. if he had four brothers and four sisters, it still would not be a fantastic statement. Now if he had two unicorns and three half elf siblings, that’s a different matter.

      • Anonymous

        If it’s impossible then there’s no chance of it happening, in gospel accounts Jesus did have 6+ siblings so the average wouldn’t be relevant.

        On probabilities: the issue is how likely it is that someone existed without any other evidence confirming it. So yes we know Josephus existed in other ways, but if I claimed that a literate man named Phejosus existed [lets assume that this was a typical name for the time] then such a person would be suspect because of how rare literate people were.

        That’s why the President (and usually lottery examples) are misleading, there is a virtually 100% chance that someone will win the lottery or become President but if someone _claimed_ they became President or won the lottery, then without the proper evidence we would be skeptical of the claim.

        • arcseconds

          You seriously think it’s impossible to have 7 siblings?

          On the basis of what? I looked at the book by Meyers cited, and it mentions that number, in passing, referring to another study. It explicitly says it’s an approximation, so we’d expect some kind of margin of error on that. I hardly think this kind of thing has the status of a natural law.

          Plus, and this seems an important point, the estimate is for numbers of people in a household compound. That’s different from the number of children one might have. If the children move out, the size of the household decreases.

          The study by Zorn also seems to largely talk about household size. We can’t just go from household size to number of possible children without making a lot of additional assumptions.

          Also, in that study by Zorn, there is this:

          “In Iron Age Tell en-Nasbeh a “typical” family size, assuming an average of 4.5, and a standard deviation of 2.5, would range between 2 and 7. Any population estimate based on average dwelling -floor area should thus fit a family size of 2 to 7. Of course, sometimes housese had to accomodate more inhabitants than they were meant for, while others may have been more sparsely occupied. ” (p 37.)

          (I included the following to sentences because they make it fairly clear that it’s household size being discussed here)

          7 is one standard deviation from 4.5, so this isn’t telling us that the maximum possible is 7. This fits with the idea that 9 is an unlikely size for a household, but hardly establishes it as impossible. If it were a normal distribution (which I doubt is a good assumption), then there would be about a 10% chance of a household size of 9 or more!

          Also, also, check out table 4 in the Zorn paper. One site in Iran has a mean family size of 8.4, and a standard deviation of 4.9! In that village, 9 would be typical!

          Can we dispense with this idea that a single paper giving an estimate for a society 1000 years prior somehow has the status of a natural law, now?

          • MattB

            I’m not a platonist. In fact, that’s why you were trying to advocate. Platonists think that Morals are some independent attribute or quality that exist apart from human beings. Quite frankly, I don’t see how that makes sense. What does it mean to say that morals exist independently? Morals are attributes of personhood. It only makes sense if you posit a moral law giver(God) who is that personhood and is a maximally great being that is the definition of what good is.

          • arcseconds

            Why are you replying here? did you reply to the wrong comment?

          • MattB

            haha no. I was just too lazy to go to the other page where this conv. was held and decide to click on your name and read what you wrote, and then posted here:)


          • Anonymous

            Meyers’ figure wasn’t just an approximation, it was the *highest* approximation possible for not taking into account things like infant mortality. This means the error bars only go down, not up.

            I think Zorn’s point was for temporary guests, I’d be odd to “sometimes…accommodate” for new children well above what was possible when the risk was starvation.

            None of the things listed are normal distributions, that’s when the average is zero and the SD is 1.

            Because Iran had the same economy and population as Israel?

            There are multiple papers used here, and they all point to a terminal limit of 7.

          • arcseconds

            I think you had better check your understanding of normal distributions:


            You can see quite clearly there that the normal distribution is a function of the mean and standard deviation.

            A couple of lines below the equation, there is this:

            If μ = 0 and σ = 1, the distribution is called the standard normal distribution or the unit normal distribution, and a random variable with that distribution is a standard normal deviate.

            You’ve conflated ‘normal distribution’ and ‘standard normal distribution’.

            The question remains why throwaawayaccount1 thinks family sizes are normally distributed. Purely on the basis of eyeballing the graphs of real family size distributions they do not look normal at all, they are definitely skewed. I would say log normal would be a better approximation.

            I think the estimates on household sizes in the Zorn paper are based in a large part on the size of the households. I suspect the difference really is just that the houses at the site in Iran are larger.

            But anyway, if you think ancient Iran has a completely different economy and population from ancient Israel, that rather speaks against this assumption that all ancient economies are exactly the same, and therefore an estimate from 1000 years before about maximum household sizes can just be assumed to hold true for Israel throughout its history, doesn’t it? Which one is it, are they all the same, or do they vary?

            The papers do not all support that as the upper limit, even for household sizes.

          • Anonymous

            I think he/she cites pre-industrial US populations as being normally distributed but yes we don’t know for certain how ancient Israel was distributed making it hard for a probability argument (and yes I was mistaken with standard normal).

            It’s not that all the economies were the same, only that none of them had growth. Their initial conditions remained but were different around the world.

            Aside from Meyers, it seems like Zorn’s exceptions was for guests making 7 the limit for families.

          • arcseconds

            Why do you keep saying that all the papers support an upper limit to 7 for households? They clearly do not, Zorn’s figures nowhere suggest this, as far as I can see, (you can’t determine an upper limit for anything given just a mean and standard deviation) and in many instances suggest that 7 or 9 would not be all that uncommon, as I have just indicated.

            Even throwawayaccount1’s normal curves, based as they are on unjustified assumptions, do not show 9 to be an impossible size for a household.

            Also, as I have already stated, all of this is quite beside the point, as all of the papers are discussing household sizes. They sometimes say ‘family size’, but it’s clear they’re talking about people living under one roof. None of this is relevant to the number of children a person could have, unless you can show that adult children stayed with their families for ever.

            (Here’s a thought: maybe some of Jesus’s brothers were illegitimate, and therefore lived in different houses! That’s not at all an unlikely background probability.)

          • There is also a very ancient tradition that Mary was Joseph’s second wife, and so the possibility of a larger family than a single marriage would produce, with no more than seven living at home at any one time, also presents itself as an option to consider.

          • arcseconds

            Yes, you’ve mentioned that before. I had thought of mentioning that… even just on the discussion of how many children in general a person could have, it suggest that even if there really was some natural law absolutely preventing 8 people in a household (I still can’t get over what an absurd suggestion this is), then having 7 children might still not be terribly uncommon.

            Dying in childbirth was not uncommon, for example. A man might, say, marry at 20, have three kids, wife dies in childbirth at 30, man remarries four years later to a younger woman, has another couple of kids in the next 5 years, by which time the oldest is 19 and leaves home. Family size is now 6, close to the 5.5 average that was mentioned earlier. The next two siblings might leave home at the same rate as another two children are born.

            Nothing seems impossible or even particularly strange about this whatsoever, except for the lack of infant mortality, but there’s plenty of time there to stick in a few kids that die…

          • And did you see that, on the blog he linked to, he basically argued that, if something characterized roughly 5,000 people in every million, then if a text says something of that sort about a person, we should assume they didn’t exist?

            I think the most frustrating thing is that our anonymous commenter seems not to have ever read the passage in the Gospels he is referring to, which makes little or no sense if it is contrasting metaphorical siblings with metaphorical siblings.

            But as always, I am grateful when a mythicist takes the time to illustrate the nature of the apologetics used to “support” it.

          • arcseconds

            I did rather enjoy the idea that if 6 brothers are unlikely, then it must be the case that all of them are metaphorical.

            It’s simply not possible for someone to have miscounted, or gotten two different sources confused somehow, or for two brothers to have been invented to make the number up to 7. Nosireebob. We could have believed 3 brothers, but if it says there were 6, there must in fact be none.

          • Anonymous

            The Mark passage was clearly not about a second wife though.

          • On the contrary, the reference to Jesus as Mary’s son rather than Joseph’s could indicate precisely that.

          • arcseconds

            Also, admitting the economies were not the same is basically admitting that you can’t use data from 10 centuries prior with any degree of certainty.

            Plus, I’m not sure where you get the idea that Zorn’s ‘exceptions’ (nowhere does Zorn appear to think that household sizes can’t go above 7) are including guests. This looks like special pleading to save your idée fixe, to me.

          • Anonymous

            Not at all, differences in the size of the economies and their growth are two different things. To say that because economies varied (given different natural conditions) is a complete non-sequitur to how much those different starting points GREW. We know that Israeli settlements (when there was a limit) did not grow into Roman times.

            If you’re changing your standard to “we need a consensus for a limit or else we don’t know” then that’s clearly absurd.

            As it stands, Meyers said the limit was 7, Zorn (from what we can tell) said that _families_ got as big as 7, the two studies, which measure different things, agree with each other.

          • arcseconds

            Perhaps we don’t need consensus, but I’m certainly not going to trust the figure that one paper gives as an estimate giving an absolute maximum limit for household sizes for all time that can never be breached ever (which almost certainly was not how the author intended this to be understood, anyway). Do you normally put this much faith in the figure one paper gives? You do know that most conclusions in papers in academia are wrong, right?

            And, as I have already explained, although Zorn uses the phrase ‘family sizes’, what he actually is talking about is people living in a single household.

            Look at table 1 in Zorn. It explicitly gives ‘number of people per household’ as the numbers. It is those numbers that are used for the 4 – 5.5 estimate in p33a (he cites Watson, Kramer and Finkelstein, and those figures are summarized in the table).

            So all of this is kind of beside the point. Household sizes are only indirectly related to how many children an individual can procreate.

            I have already shown that Zorn does not support a maximum of 7. The standard deviation he gives on p37 is 2.5, which easily rules in 9 as a potential family size. Why are you continuing to say he does support this? Is this your normal strategy of winning arguments, keep repeating your claims no matter what your interlocutor says? If you keep repeating this without giving an argument for it, I’m going to have to do something… like commence mocking you, maybe. You have been warned.

          • Anonymous

            I don’t know if I would say *most* papers are wrong, but Zorn explicitly gives 2-7 as the capacity regardless of the SD. What Zorn is thus saying is that 5% of people (since 9 is about two standard deviations away) lived above capacity, while Meyers is saying 0% of people did.

            You can look at it as a 0-5% chance of it happening or as 95-100% chance of it not happening. I think it makes much more sense to think of it as the latter since we don’t have any other good information confirming it.

          • arcseconds

            More of your idée fixe. Zorn gives 2-7 as a 2 SD range in which 68% of the households will fall, assuming a normal distribution. He is saying that 2-7 should fit into the average-sized house. This is not saying that all houses can only fit 7 people, or that larger households must be cramming the people in to over capacity! They’d just be living in the larger houses!

            I see he is indeed making the assumption of a normal distribution. He is not entitled to just make the assumption, and I can’t see any defence he gives of this. It also seems unlikely to me, I’d expect household sizes to be more log-normal, with a minimum size, a mode around the ‘modest dwelling’ size, and a long tail of larger and larger households getting less and less unlikely, pulling the mean above the mode. This is roughly what you see if you look at actual graphs of family sizes.

            But assuming a normal distribution (partly because I can’t find any good log-normal distribution calculators out there), there is about a 16% chance that a household at that particular site would have more than 7 people. What kind of absolute upper limit has 16% of cases falling above it?

            Also, in a normal distribution with a mean of 4.5 and a standard deviation of 2.5, it’s 9.5 that’s 2 standard deviations away from the mean, not 9. This makes a difference: the chances of a given household being greater than 8 (meaning 9 or more, as you can’t have a fractional person) is actually 9.1%.

            What justifies you in changing a 5% chance to a 0-5% chance, anyway? Can I just arbitrarily add 5% onto a 9% chance and make it 14% too?

            And what do you mean ’95-100% chance of it not happening’? If the distribution is normal, then there definitely will be households of 9 or more people. There will be lots! We’d expect there to be 20 such families in Tell en-Nasbeh alone.

            If there’s a 10% chance of something happening, are you just going to disbelieve all reports of it happening as though it were impossible? If I tell you I have a friend who’s 190m tall, are you going to say “no way, mate, the chance of that happening is less than 10%, so you have no such friend?”

          • Anonymous

            I went back and re-checked the numbers and it’s worse, when you run an average of 4.5, SD of 2.5 and our value of 9+ (which is better because it allows for more possibilities), you get a 3.5% chance.


            “I see he is indeed making the assumption of a normal distribution. He is not entitled to just make the assumption, and I can’t see any defence he gives of this.”

            Well I’d hate to say it but you’re not one who published the paper, I’m assuming most family populations end up being normally distributed.

            “This is not saying that all houses can only fit 7 people, or that larger households must be cramming the people in to over capacity!”

            Yes it does, he says

            “Any population estimate based on average dwelling -floor area should thus fit a family size of 2 to 7. Of course, sometimes houses had to accommodate more inhabitants than they were meant for, while others may have been more sparsely occupied.”

            During those times, this would mean being economically above capacity.

            By his estimates 3.5% of people went above capacity while Meyers is assuming 0% did. Which means 0-3.5%.

            I don’t think I have explain why we should assume the more likely event happened (the 96.5-100% chance that there was no such family) rather than the unlikely event.

          • If only one person in a hundred in the U.S. failed to graduate high school, given the millions who would fall into that category, would you doubt someone who claimed not to have graduated from high school? I still don’t think you are bothering to do the math.

          • Anonymous

            If you didn’t know anything else (and had no way to check) yes you would doubt it, that’s how statistics work.

            A better example is if someone claimed to be one of the top 1% richest people in the US, and you had no way of knowing. Obviously you would doubt such a claim.

            We go with what’s more likely, and in a comparison of a 97-100% chance or a 0-4% chance, it’s clear which one is more likely to be true.

          • Thank you for clarifying how wrong I have been in how I have proceeded in our conversation. Mythicists are very rare and I had no business believing that you are one.

          • Anonymous

            Hilarious although I’d venture to guess that the reference class on this blog isn’t randomly selected people.

            If you go to a top 1% richest people blog, then yeah the chances would increase quite a bit for that too.

          • arcseconds

            The question is not which is more likely, the question is what probability should you give that something with a 4% background chance has happened given you have a report about it.

            If you stick with the 4%, then what you’re telling us is that you never give any credence in any report. That seems absurdly sceptical, don’t you think?

            Here is the data for US household sizes in 2006:
            Household size : %
            1: 27%
            2: 33%
            3: 17%
            4: 14%
            5: 6%
            6: 2%
            7+: 1%
            ( source: )

            So every single possible number has a less than 50% chance of being true.

            If we apply your logic, we’ll never believe any report of any household sizes. Someone who claims to be living in a household of 2 has a 77% chance of lying, according to you!

            Hopefully that makes the problem clear.

            What normal people would do, and what I think you too would ordinarily do if you weren’t trying to be all Bayesian rational and failing miserably at it, would be to immediately upgrade the probability close to 1 as soon as receiving the report, on the basis that people aren’t likely to lie or be wrong about how many people live in their house.

          • Anonymous

            No that’s the probability someone telling you they are in a class is lying (which does not apply in the case of Jesus).

            And it’s not in a vacuum, the alternative is tested and is significantly more likely (the blogger recently updated it to include the mythicist probability). It’s 0-4% vs. 27%.


            “In the phrase you quote, the only reason why the range is given as 2-7 is that it’s 1 SD away from the mean. That’s it. There’s no reason to suppose that no house can fit more than 7 people.”

            I think you’re mixing things up: there’s a capacity/limit for survival but this doesn’t necessarily mean it’s an impossibility. Zorn is saying the two are different while Meyers is saying they’re the same. That’s why the probability is 0-4% including both figures for how many houses went above the limit.

            At any rate, if we’re making a probabilistic argument (choosing which is more likely) then I which do you think we should go with?

          • arcseconds

            Honestly, you have become so obsessed with the number 7 that you’re allowing this to distort the meaning of everything you read. Because you found it mentioned as an upper limit in one source, you’re now determined to make it into some kind of a limit in every source you read.

            Zorn’s data shows that 7 as an upper limit is untenable. So at best Meyers’ upper limit of 7 is only valid for Iron I.

            You admit this when you allow for 3% of households to have over 9 people! (I think your earlier figure of 5% is better, I admit that 9% is too high, I’ll explain in a bit). But you still want to read in 7 as some kind of a limit, even if it’s not the maximum limit, so now you’re misreading this passage to say that there’s no way anyone ever built houses that could have more than 7 people in them, but sometimes they squeezed more than 7 people in.

            The context here is that Zorn is about to compare ethnographic data with the floor area of houses and estimates of what floor area supports how many people. The idea is that in theory the floor area of the houses should be plausible given the estimates of household sizes. An average house should fit an average household. As he says on p35,

            “If the resulting figure is absurdly high or low, there is something wrong with either the ration used or the average family size used, possibly both”

            In the phrase you quote, the only reason why the range is given as 2-7 is that it’s 1 SD away from the mean. That’s it. There’s no reason to suppose that no house can fit more than 7 people. All that’s maintained is that an average house should be able to fit between 2 and 7 people. If it can’t, then there’s something wrong.

            Look at table 3. That is the building sample from Tell en-Nasbeh that he is working with. There are a range of building sizes. The mean is 36.5m² (from the ‘less special’ figures). The largest has 75.5m². Do you think that building also can only house 7 people? If we ingore that one as an outlier (might not even be a house) the next largest is 48.2m². That’s about a third larger than average. If 36.5m² can squeeze in 7, then 48.2 can certainly house 9.

            He’s not sure whether or not the houses would have had second stories. He’s actually happy to accept either that they did or they didn’t, both don’t end up with absurdly large or small figures for floor area per person.

            The point here is not that the largest house would have had 9 people in it. The point is that there’s nothing here that implies that there’s some kind of limit of 7 people for every single house. Only that if it winds up that the average house is massively spacious for 7 people, either the 2-7 estimate for most households is wrong, or the floor area estimate per person is wrong, at least for this site.

            ‘Of course sometimes houses had to accommodate more inhabitants than they were meant for, while others may have been more sparsely occupied’ is just a caveat pointing out that he’s dealing with averages, not saying that every house can only ever have 2-7 people. Some of those average houses may have housed 9 people sometimes, but we might expect that larger houses than average would house more people than average.

            Seriously, you are completely misunderstanding this paper if you think that it shows that no house can have more than 7 people.

            As far as normal distributions go, it’s a common mistake to assume normal distributions when there’s no reason to. If he has reasons for thinking that it should be normal, he should mention this. As I have already shown, family sizes are not normally distributed. . One would not expect them to be, either: people have to go through 1 child to get to 3, so you’d expect at a point in time there will be more smaller families below the mean just because they’re younger families that might grow, and the mean to be bought up above the median by a few very large families.

            The distribution of the houses given in Table 3 actually, on initial inspection, doesn’t look normal either. It’s very definitely skewed: 60% of the houses are below the mean.

            One problem with calculating probabilities using a normal distribution for household sizes is that any real household can only have whole numbers of children. I would say this is another reason to be suspicious of normal distributions being used here, particularly when the numbers involved are small (a normal distribution might be a better approximation to a discrete distribution if we were talking about ranges over 100 or more).

            All Zorn is doing is getting a rough estimate for how large an average household can be. His analysis is so seat-of-the-pants that maybe this is fine (he’s really just doing ballpark-figure stuff, and I think that’s all anyone could do on this data).

            Anyway, when using a normal distribution to approximate discrete data, it’s usual to take 8.5-9.5 to represent ‘9’. So I accept your earlier figure of 5% for a 9+ family. It was incorrect of me to use >8. Have you any reason to use >9 rather than >8.5 to represent ‘9 or more’?

        • Michael Wilson

          you write, “Some background: a study of Israel settlements said that it was economically impossible for an average family to be larger than seven.”

          In the set 2-2-0-8 it is impossible for the average to be higher than 3, the average of those for numbers ( Add the four numbers that make the set, then divide by the number of numbers making up the set, 4). But yet an individual number in the set is 8. See how averages work?

          • arcseconds

            I think the problem is not that Anonymous doesn’t understand averages, but that they think the figure mentioned by Meyers ‘For the nuclear families (brothers with their spouses and offspring) that were present in family compounds a maximum of seven persons can be estimated…’ shows that it’s completely impossible for a couple to have more than 7 children.

          • Anonymous

            Hmm here’s an interesting exercise: find where Meyers uses the average to determine the limit for large a family can get (here’s a hint: she doesn’t).

            Nor do I ever argue that there was a limit based on the average, I’m assuming you’re conflating the argument about probabilities (which does use an average) and the argument for a limit (which does not).

            Plus it’s a bit ironic of you to accuse someone of not understanding statistics when you yourself didn’t know what a normal distribution was.

          • arcseconds

            Are you replying to the right person? Michael Wilson thinks the argument on maximum family size is based on averages, not me.

          • Anonymous

            Oh woops I literally missed the “not”, but yes my bad that was to Michael.

          • Michael Wilson

            So you dont think it impossible that Jesus had 6 siblings, only that it would be less than 1% chance that he would have. I could agree with that. Growth rates in America are returning to pre industrial levels, so our rate for 7+ person homes might not be a bad assumption, though at the time I think large famillies were in demand, so you would make them if you had the resources and health. That being said, if it were a figure of one familly in 300 every other village would have such a family. These would cluster toward wealthier families and those families would have been lucky regarding birth rate, survival and mortality rate. It ought not to be shocking Jesus came from a prosperous familly. Most great figures do.

          • Anonymous

            The question is whether the gospels wrote that Jesus had a biological family, and I’d say it ranges from impossible to less than 1% which without any evidence is basically a rejected hypothesis.

            McGrath has also proposed that the gospel was wrong on this point which I think is less likely and hurts the historicist position more for making Mark unreliable to determining family relations.

          • Michael Wilson

            So if I say X is part of 1% of population y, all else equal the chance of that being accurate is 1%? Statistically, that is rare, not impossible, but it doesn’t affect the reliability of the statement. Independent evidence suggest at least 2 brothers. I can neither confirm or deny more.

          • arcseconds

            Yes, in Bayesian terms (i.e. using probability in a Bayesian sense of personal belief-weightings, not necessarily using Bayes’s theorem) we would expect a report that we think is even mildly trustworthy (more likely to be right than wrong) to increase our probability of this quite steeply.

            The background probability still has a role to play, though. An incredibly unlikely event will need a very reliable source to back it up. It’s not very likely that Jesus had 50 siblings, so unless we had absolutely fantastic sources for this, we’d probably be disinclined to believe it (even if we thought the source was actually usually pretty reliable).

            (EDIT: I actually did try plugging this into Bayes’s Theorem, but I got stuck on the usual stuff like deciding on reference classes and pulling probabilities out of my arse. How do I know what the chances of a text mentioning six siblings if there aren’t, in fact, six siblings, is? Should that be any text, or just texts mentioning siblings, or what? But anyway, FWIW on assumptions that the text is unlikely to give us the right answer, but six is even less likely to be mentioned if there weren’t in fact six, it’s easy to go from a 1% background chance of six brothers to a 5%. 1 in 20 ain’t too bad. my faith in this calculation representing the actual chance of there being six brothers is not high, but it does illustrate the point of even unreliable evidence raising probabilities markedly)

          • Michael Wilson

            Yes, often I find that ancient history isn’t suitable for precise statistical analysis. We have go with ball park figures. The odds of Jesus being part of a large familly aren’t high, but neither are the odds of Peter being a fisherman or any number of things. But they arent so rare as to makes doubt the veracity. But if you said 3 magi from the east visited you at your birth, I would be very suspicious.

          • arcseconds

            I don’t think it’s usually that suitable for precise statistical analysis either, but I do agree somewhat with the sentiment expressed by someone discussing the matter with Ian a few days ago that Bayes’s theorem, and formal probabalistic reasoning more generally, has some potential for structuring the reasoning process. That’s not to say that you necessarily just accept the number that’s spat out of some calculation – you might say “well, that’s absurd” which might mean you adjust some of the inputs, have a look at your reference classes, etc. (one man’s modus ponens is another’s modus tollens). I think one could learn something from this process.

            For example, you might find that you’re not as certain about some things as you thought you were, or maybe you’re even more certain. Another interesting possibility it that it might make it transparent that someone could differ from you on the inputs but not by a huge amount, but reach very different conclusions, in a perfectly sound manner.

            Of course, to do this properly is quite difficult, maybe too difficult. You’d need to consider ranges of probabilities, and also it’s really hard to give a numeric value to how trustworthy sources are, say, on the number of family members, particularly as we don’t usually have non-textual evidence for that.

          • Michael Wilson

            It may have use, but I find data for antiquity so sparse that precise calculation is impossible. I wouldn’t want to sell insurance to the past. Looking into this though has shown me how unusual Jesus family acording to Mark is. I hadnt thought of that and I do think if true, it would point to a much higher chance of Jesus belonging to the top half of incomes than I had allowed for before. The notion that it suggest that Jesus is myth of Mark mistaken is a bit more than is warrented though.

          • arcseconds

            Precision isn’t possible, no, but Bayesian probabalistic reasoning is promoted by its supporters as a formalisation of inductive reasoning generally, so in principle any such reasoning (well, any such reasoning that’s any good) ought to be able to be given a Bayesian formulation.

            And it’s supporters aren’t limited to just Carrier. There are a number of philosophers of science in good standing that support this notion.

            Of course, even if inductive reasoning is in principle formalisable as Bayesian statistical reasoning and receives it’s ultimate justification from that fact, that doesn’t mean in practical cases, especially in areas like history where ulimately there are many judgement calls, it’s really feasible.

            Anyway, you seem to be far more impressed with this analysis than I am. I wonder why that is? I don’t think I’m left with any firm conclusions whatsoever, for reasons I’ve already stated several times. It could well be that Jesus’s household was pretty average, but Joseph married twice and had children over a long period of time. None of the household data gives us any information about how probable that kind of thing would be, but I don’t see any reason to think it all that uncommon.

            Also, it’s entirely possible that Mark’s list of brothers isn’t correct, for whatever reason. Maybe some of them were cousins, or it was a later attempt to establish some important early Christians with a close link to Jesus, the importance of which is now lost, or anything.

            I suppose if pressed I’d probably kind of agree with you that a more well-off than usual background is probably the most likely option, and it does mesh with a large family size, but if you were already fairly sure that his background was well-off, the evidence here shouldn’t make you a lot more certain.

          • Michael Wilson

            It is possible Mark is mistaken, but we know from paul that Jesus did at least have multiple brothers. It is only one piece of data to try and appraise the wealth of Jesus’ familly. I do think that data cuts against a veiw like Crossan’s that Jesus came from a family of manual laborers (though Jesus himself was called a tekton, not his father, so we really don’t know what his father did. Jesus himself may have pursued a differnt vocation, possibly a manual laborer) . I very much doubt 7+ would have been usual. If you look into it, you will see that population growth was very slow before modernity and tha means big families were uncommon. The culprite is high infant mortality and scarcity of food. Getting 7 kids to all be alive at the same time would be quite an achievement before modernity, and it would be harder if you were poor.

          • So basically you are happy to distort what I said, lie about what the Gospels say, misrepresent the articles you cited as well as those I and others have shared with you, and are going to plow ahead despite having the relevant math explained to you, insisting that Iran must have been different but that Palestine a thousand years later must have been exactly the same?

            If you keep this up, it only serves to confirm the impression that many of us have of mythicism and of the tactics used by its proponents.

          • Anonymous

            What I am seeing is people mix up an argument for probability (which uses averages) and one about limits (which does not) and one about growth (not economic starting positions). So I’d say it’s the historicists who either don’t understand the argument or are intentionally conflating it.

            How have I distorted anything you said? You said that maybe the gospel made Jesus’s family bigger for some symbolic reasons.

            You might not think it makes it unreliable on this point, but that is what logically follows.

          • So far you seem not to have understood the nature of the region’s political, economic, and social history, and have not understood the passage in the Gospels that you are making claims about. If you have not misrepresented, you have at least misunderstood much that has been presented in this discussion.

          • Anonymous

            Sorry but when the conversation goes:
            “You don’t understand.”
            “I mentioned x,y,z, and people thought q.”
            “You don’t understand.”
            It’s clear which side doesn’t want to engage the issue.

          • You just keep saying that one article says something that supports your conclusions about a different time and context, and you keep showing that you don’t even understand the passage in the Gospel that you claim to be trying to interpret. If you imagine that the problem is others not engaging the issue, then you are delusional.

          • Anonymous

            No there’s multiple articles confirming the conclusion and as I’ve been saying for the nth time: The economic context is the same.

            Economies did not grow four thousands of years until 1800s, you have not responded to this fact once.

          • The economic context before Israel was a nation is not the same as when it was. The economic context was not the same in the time of Omri as it was in the time of Jehu. We have evidence of the economic changes in the region, and your constant assertions that things never varied just shows that you don’t know the history. It doesn’t change the facts. And houses did not all remain the same size, allegedly tightly designed to prevent anyone from ever having a larger family than 7. Please stop pretending you do not understand that the evidence does not support your claims. It has been pointed out enough times that you cannot actually have failed to understand at this stage.

          • Anonymous

            No we don’t, to quote the economic historian Gregory Clark, “the average person in the world of 1800 was no better off than the average person of 100,000 BC.” This is what social science says about the history of the economy. Here’s a picture to show the data on it:


            Things like culture might have changed, but the economy did not grow.

          • Of course things for the average person remained the same. Even after the industrial revolution, it was the drop in prices of some goods that benefited the average worker. Their wages themselves did not increase.

            But if you are suggesting that a farmer’s life was the same during a good harvest as during a locust infestation, or that a mason living in the vicinity of Samaria had life the same before or during the reign of Omri, then you are merely showing once again that you do not understand what is being spoken about or are assuming that every single person and every single year was average.

          • Anonymous

            I’m not suggesting that, the relevant question is the economic capacity to support more children, which did not change until the 1800s.

            And again for the nth time, this does not mean that the entire world was exactly the same, even on that chart there’s little differences between non-growing economies but the fact that none of them grew did stay the same.

          • arcseconds

            Perhaps you are confused about what the historicist position is. The historicist position is merely that there was a figure that meets a fairly minimal description existed. Usually this not much more than:
            *) their name was Yeshua or similar
            *) they had an active ministry around 30 AD
            *) they were crucified
            *) they were casually responsible for Christianity in the way a charismatic religious figure would normally be considered to be responsible for the founding of their sect, for example it’s this Yeshua that gave rise to the stories of the crucifixion in the Gospels, and it’s this Yeshua that Paul attributes all sorts of things to in his letters.

            The Gospels can be wrong about all sorts of things yet still have this as the best explanation for much of what we do find there, and in the other sources we have.

            Practically nothing rides on Jesus having a particular number of brothers.

          • Anonymous

            Except that the minimal position also relies on James being the biological brother of Jesus and thus the gospels are talking about a biological family. I haven’t heard anyone openly say they can have a case without that.

            What I am showing is that either the family was metaphorical or the gospels were wrong on James being a biological brother.

          • Can you kindly explain how the passage in the Gospels that you have been referring to could make sense if it was contrasting metaphorical siblings with other metaphorical siblings?

          • Anonymous

            There’s no contrast because the people in the town (at least in that instance) were not followers. They mentioned his other followers and Jesus left after they “took offense” to his preaching.

          • I am not following your interpretation of the passage. Can you please explain exactly how you understanding it and your identification of various people mentioned in it? I assume that you don’t think that it was his followers who thought he was out of his mind…

          • Anonymous

            There’s (1) the people of Nazareth in that instance, who are not followers and (2) there’s Jesus’s “siblings” who are followers.

            (1) is mentioning (2), there’s no “contrast” or confusion over followers talking about other followers.

          • So you have not understood the story. That would explain why your view of this matter is so bizarre.

          • Anonymous

            And what do I not understand? Were townspeople who “took offense” supposed to be followers in that instance?

          • Here is the relevant passage from Mark 3:

            31 Then Jesus’ mother and brothers arrived.(AD) Standing outside, they sent someone in to call him. 32 A crowd was sitting around him, and they told him, “Your mother and brothers are outside looking for you.”

            33 “Who are my mother and my brothers?” he asked.

            34 Then he looked at those seated in a circle around him and said, “Here are my mother and my brothers! 35 Whoever does God’s will is my brother and sister and mother.”

            How do you explain the meaning of the details of this passage on your view?

          • Anonymous

            Ironically it could be Jesus telling the people not to confuse the “brothers” for biological brothers but rather symbolic ones. (as in, “they aren’t my biological brothers, my brothers are anyone who follows god” etc.)

            Although I agree that by including Mary and “mother” it makes that a bit convoluted. I’d say it’s 50/50 either way in that instance but including all that we know, the term has a higher chance for symbolic.

          • It could mean anything at all, if you ignore the actual wording and the context of the Gospel in which the words are found. But I am looking for an interpretation that takes seriously those things you are happy to ignore and never discuss or consider important.

            Is the 50/50 odds just something you pulled out of thin air? And whatever are you referring to by “all that we know” when apparently you didn’t know this familiar text until I drew it to your attention?

          • arcseconds

            I think anonymous was referring to Matthew 13:55-56. That’s where we get at least six siblings from, and what the original site referenced. I think the suggestion is that they’re followers, not literal brothers. Not that this makes any more sense…

          • He seems to be using that tried and true method of Christian apologetics. Start with a text that you can interpret as meaning X when taken out of context. Then, when counterevidence is encountered, insist that it cannot mean Y because you have already established that X.

          • Anonymous

            So are you going to give the reasoning or just keep saying “you don’t understand”?

            Assuming all symbolism, Jesus contradicts the more common understanding of “brothers” for fictive kinship. It’s not like he randomly brings it up, it’s a reminder not to confuse the two (albeit not perfectly because he includes Mary thus making it not fully in favor of either side, ie 50/50).

            “All that we know” includes Paul’s letter’s, frequencies of ancient names etc.

          • You list evidence that supports mainstream historians’ conclusions, but then suggest that the evidence could go either way. You aren’t making any sense. Of course, if you assume the meaning of the text is symbolic, you can always crowbar evidence into your chosen “conclusions.” But that isn’t scholarship.

          • Anonymous

            I’m saying this particular passage (Mark 3), can go either way but that the total evidence (gospels where Jesus calls hundreds of followers “brothers”, Pauls letters with name frequencies etc.) point to symbolism.

            As a complete sidenote, in OHJ Carrier also tests Jesus’s historicist probability using a class of Jewish prophets and comes out with a low result. I’m just saying because every reviewer criticizing the book has apparently missed it so far and I think you mentioned that as the relevant class to use.

          • A Bayesian analysis merely quantifies the probabilities, but it will not correct for places where Carrier’s assessment of what is probable or understanding of the data is open to challenge. You can be mathematically rigorous and still be wrong if the numbers you are plugging into your equation are dubious.

            You seem not to understand that the fact that there are groups, religious and otherwise, who use “brother” metaphorically, does not preclude people from having literal siblings. You also seem not to understand that the grammar used in the relevant texts often indicates that one or more literal siblings is in view, who is/are being contrasted with metaphorical brothers/Christians. Don’t you think that studying the texts ought to be your first step? Again, judging probabilities based on a dubious or partial grasp of the evidence is not going to lead to a persuasive historical construction, even if you use the right equation. Garbage in, garbage out.

          • Anonymous

            Concern over a low prior probability can be offset by good posterior probability. That is, even though no one in Jesus’s class of Rank-Raglan heroes existed (Zeus, Moses etc. as priors), good posterior evidence can offset that by showing he did.

            Anyways all I was saying was not to miss Carrier’s alternative class of Jewish preachers like all critics so far did.

            It’s pretty clear that thinking more evidence points to one side =/= me thinking the other side is impossible, unless you were making a blatant strawman.

            And that is what we were doing for the text, I’m asking for _your_ arguments because you’ve presented none so far.

          • Comparing Jesus to figures who were gods, or about whom our earliest sources are hundreds of years after the figure is supposed to have lived – and whom we don’t know were not based in historical figures – shows that you amd/or Carrier are not even really taking this seriously.

            I have been blogging about this subject for almost a decade. And historians have been writing about it for MUCH longer than that. If you have not found my arguments, I suggest that you begin taking this subject a bit more seriously and look a little harder.

            Here is a good place to start:

          • Anonymous

            They aren’t gods necessarily (Moses was not a god), they’re Rank-Raglan heroes. On the scale, Jesus (just from Mark) is in a class people where none are historical. But that and the Jewish prophets class are of course all in the book.

            Well historians have certainly used faulty methods and assumptions for decades but where did you lay out the full case for biological siblings?

          • I’ve not read Carrier’s book. Is the Rank-Raglan list of heroes Carrier uses identical to the one in Raglan’s original work? Rank, from what I remember, wasn’t remotely interested in historical questions.

          • Anonymous

            It definitely uses the same methodology but the people aren’t necessarily the same that Raglan used. It looks at high scoring people on the scale in the Mediterranean around the time.

            The point is Jesus is the same class of people as Zeus and Moses but is the only one considered historical.

          • Does Carrier mention Cyrus the great or Mithridates VI? I’d be interested to know how he avoids lumping them in with fictitious characters like Zeus or Jesus.

          • Creating a category that includes both gods and legendary humans the stories about whom may ultimately derive from a historical individual, and then placing Jesus in it, does nothing whatsoever that is useful in relation to questions of historicity.

          • Anonymous

            Are you saying Rank-Raglan is a bad methodology for determining historicity because that’s essentially what you’re describing.

            Also do you know around when your review of OHJ is going to be finished?

          • I will be interested to see whether Carrier makes better or even different use of the scale than Price has.

          • arcseconds

            Even if it’s true that the number of brothers is given as 7 for metaphorical purposes, (and we’ve already seen that you have practically no basis for claiming this) then why are the only options ‘all brothers are real’ or ‘all brothers are metaphorical’?

            Why is it not a possibility that Jesus had two brothers, say (these are attested to in other texts), and the number was expanded to 7 for whatever metaphorical reasons?

          • Anonymous

            Because the former doesn’t make any sense while the latter does.

            There would be no reason to just portray a large biological family when having an entirely metaphorical family makes sense, especially in the context of Jesus preaching to dozens of “brothers” in other gospels.

          • arcseconds

            Er, what?

            I asked you why these are the only two options. You don’t answer that question, but just continue to assume that they are the only two options, and justify why you think the latter option is the correct one.

            It seems really, really hard for you to revise anything once you’ve decided to assume it, isn’t it? To the point where you don’t even understand what people are asking you, and continuing to repeat claims when they’ve been dismissed.

            So, I’ll ask again: why assume that these are the only options? If it’s so important to Mark to have 7 brothers for some reason, why isn’t it possible, or even quite likely, that Jesus had two brothers, James and Judas, and the other five were invented?

            On what you do say… it doesn’t make sense to mention six brothers if Jesus actually had six brothers? What on earth are you saying here? It almost sounds as if you just expect Mark to lie outrageously: if there were six brothers it wouldn’t make sense to say there were, but if there weren’t six brothers that would explain why he says there are?

          • Have you actually read the Gospels? You seem not to even be familiar with the one text about which you keep making claims.

          • Michael Wilson

            Having thought a little more on this, pinker’s medieval figure suggest that the number of middle class and higher famllies with 7 or more members may be higher than the >1% today’s stats suggest. If Jesus did come from a family of artisans rather than landless laborers, which again is common among people noticed by history, his families chance of bring large could approach the early industrial level where traditional desire for large families met sufficient food and sanitation but not much better medicine. In short, the upper classes of Jesus time may have had large families at the rate of 19th century Americans generally.

          • arcseconds

            I don’t know anything about the material conditions of Iron I (and I didn’t even know there was such a thing as Iron I until a couple of days ago, at least, outside of Iron (Fe) in its first oxidation state…), but wouldn’t there being more artisans, and more wealthy artisans, be one of the major differences between early Iron Age Palestine and 1st Century Palestine?

          • Michael Wilson

            Based on what ive read, the population and GDP of 1st c. CE Palestine was quite a bit bigger than iron I Palestine. I imagine that individual household size didnt change as much but I think the odds of their being 9+ famllies around would be much higher. Your right to, their would be more wealthy artsans who could feed large families.

          • MattB

            Plus, much of what archaeologists and historians know about Nazareth is that it was only 4 acres and that it had 50 houses. That would mean that there were probably a lot of cramped families in a tiny village. 9+ family members doesn’t sound so improbable.

          • Michael Wilson

            well, the question is whether he had 6 or more brothers and sisters so whether they were packing families into houses like cordwood doesn’t factor in, just whether a guy could have six or more brothers and sister and if he odds of that happening are so low as to cast doubt on the claim.

          • MattB

            oh, okay.

          • arcseconds

            Frankly, none of the evidence given sheds much light on how many children a person could have, and given the kinds of data exhibited, they can’t shed much light. All they can tell us is how many people can live in the same household. If there were significant incidence of children being produced over a long period of time and the older ones moving out, or significant incidence of bastardry (and one would expect both to be around to a reasonable extent), then the number of children someone could have (a male someone in the later case) could be larger.

            We would need quite different data to determine this, much more difficult to assess from physical evidence, I would think.

            Also, Zorn’s data is also on much earlier times. Expecting 1st century Palestine to look much like Palestine a thousand years earlier is obviously a huge assumption, although your figures about Nazereth suggest cramped conditions making small household sizes likely. (edit: oops, not your figures, Matt Brown’s)

          • Anonymous

            In Luke, Jesus’s family couldn’t afford a goat for passover sacrifice so from the accounts it’s clear that they were at least not a wealthy family. I don’t think the gospels thought they were poor/landless by any means but also weren’t rich.

          • I think I know what you are misremembering about Luke’s Gospel, but rather than tell you what I think you meant to write, I think it will be more fun to ask you this: where in Luke’s Gospel does it say that Jesus’ family could not afford a goat for Passover?

            While it is very unlikely that Jesus was rich, in the sense of being part of the wealthy elite, that he may have been from a family with above-average wealth in the retainer class is possible, and it is worth pointing out that Paul actually says that Jesus was rich, and made himself poor for the sake of others.

          • Michael Wilson

            Luke’s account is likey an invention, so his description of their sacrifice is suspect. It carries less authority than Mark’s notice of familly size. Of course if his familly did have 7+ offspring their may not have been much left over for luxuries so the perception that he had a modest home may be accurate.

          • MattB

            What evidence do you have to support that Jesus’ family is not biological? The Gospels clearly portary his family in a remote part of the Roman empire. In a real village. In a real time period. It would be different if the gospels said something like “Jesus’ had 100 siblings on planet Xenon and lived in a city called Zeus”.

    • arcseconds

      What justifies the assumption of a normal distribution?

      Do these look normal? I think they’re more log normal.

      Also, what justifies the assumption of a standard deviation of 1.8?

      Also, I think he’s undermined his entire point by pointing out that in th 1700s, the average was around 5, but a plurality of families were above 7. He’s already happy to assume that production has stayed exactly the same since ancient times until 1800 (!), and he’s just mentioned a distribution which is completely unlike the one he goes on to establish by assuming a normal distribution and a standard deviation of 1.8.

      Why isn’t he assuming that the distribution is like the real one that he mentions and is in the time frame that he’s happy to treat as all being exactly the same? In which case, above 7 is not just possible, but expected.

  • Nick Gotts

    The anonymous blogger says: “It’s a well held consensus that the economy has stayed the same or decreased until the mid-1800s”.
    Then no poor-to-average family before the mid 1800s should have had 7 or more children. My probable great-great-great-great-grandparents in the paternal line had 9 children between 1788 and 1806. I don’t know their exact economic circumstances, but all their known descendants for several generations for whom I do have information appear to have belonged to the English rural proletariat. Fortunately i’m descended from the fourth child, so I’m not obliged not to exist on Bayesian grounds, but I feel I’ve had a somewhat narrow squeak.

    • Anonymous

      England had both a larger economic base than Israel and was starting to industrialize around that time. I think there’s a misconception between the world being the same and growth being the same (the latter which is what’s true).

      • Nick Gotts

        Neither is remotely true, of course, but what the anonymous blogger says is quite clear, and I quoted it at the start of my comment.

  • Andrew Dowling

    Does “anyone has relevant archaeological, socioeconomic, and other perspectives to offer on this?”

    Apparently not. 🙂