After analyzing Educational Attainment + cognitive ability (EA) polygenic scores across 53 ancient cultures, the data tells a story that upends several historical assumptions. While my 2023 study revealed Rome's exceptional cognitive genetics and my 2024 paper
Good post. It is refreshing to be moving away from the whole "weak Early European Farmer" narrative that has proliferated in misinformed online circles.
Our Neolithic European ancestors were patrilineal people. Patrilineality is a kinship system whereby an individual's family membership derives from the father's direct paternal lineage.
One of the earliest examples of patrilineality comes from Neolithic Europe in 4800 BC, where two families buried in a necropolis in Gurgy, France, are both connected via the male line as determined by the structure of the burial and Y-DNA haplogroup analysis.
Further evidence suggests that patrilineality was a kinship system used by Early European Farmers of indigenous European Hunter-Gatherer origin as well. The powerful elite buried within the Neolithic passage tomb at Newgrange belonged to Y-DNA haplogroup I2a (a Western Hunter-Gatherer lineage). You mentioned in the article that Early European Farmers in Britain underwent selection for intelligence.
Despite the Newgrange elite's overall autosomal ancestry being inherited from Anatolian Farmers and roughly a quarter from the Western Hunter-Gatherers, his phenotypic traits (specifically his complexion) resembled that of his hunter-gatherer ancestors several generations before him due to being the result of an incestuous relationship. Likely as a means by his parents to preserve the phenotypic traits of their hunter-gatherer ancestors.
The system of patrilineality for cultures primarily descended from European hunter-gatherers on the direct male line continued into the Middle Bronze Age.
Recent discoveries show that the use of patrilineality by farmer-descended populations was not affected by the introduction of Western Steppe Herder ancestry into the European gene pool. This is important to note because it does away with the misconstrued idea that the Early European Farmers simply vanished or whatever people say when the Western Steppe Herders arrived. A new paper by Allentoft et al. (2024) provides an example of an individual labelled NEO792 whose overall ancestry was 85% descended from Western Steppe Herders, but his Y-DNA lineage was I2a-S2703. A Funnelbeaker Early European Farmer subclade of Western Hunter-Gatherer origin.
NEO792 was buried in a megalithic tomb, not a burial mound, and carried mt-DNA haplogroup U2e2a1, a Western Steppe Herder derived female lineage. This means that despite NEO792 indirectly inheriting the majority of his overall ancestry from Western Steppe Herders, he was likely aware of his direct paternal heritage going back to his Funnelbeaker farmer male ancestors and was given a special megalithic tomb burial in the same fashion as them.
This shows us that patrilineality was a core social attitude of Bronze Age Europeans of Hunter-Gatherer origin and that their identity, cultural customs, culture, and overall perception of themselves was unique on a local level. Further evidence suggests that this is the case based on the aforementioned paper by Allentoft et al., where a new discovery shows us that the origin of the Germanic people lay with a core population of men belonging to Y-haplogroup I1, an indigenous European hunter-gatherer lineage, with additional hunter-gatherer ancestry.
The I1 cluster's introduction of stone cist burials into Denmark by 1750 BC shows us that a unique culture had developed within Scandinavia before or around 2600 BC and that this population would have likely used patrilineality as a means of kinship identification. This system and understanding of their ancestry would also be used as an exemplification of their elite lineage moving into the Nordic Bronze Age, where we see I1 become the leading lineage of the Nordic Bronze Age elite. Not a Western Steppe Herder lineage, but a hunter-gatherer lineage indigenous to Scandinavia itself.
I believe Kevin Bird was arguing that polygenic predictions work well for Europeans but are less effective for populations like Indians or Africans, who have distinct genetic profiles. East Asian genetics also seem to be severely underrepresented, at least based on what I've seen on Google.
He suggests that accurate predictions can be made for European populations, but these break down in non-European populations due to their differing genetic makeup.
Essentially we need to understand more about the theory of mind how G expands across cultures and if G is the final concept are there more intelligence?
Can we make vr tests to measure it's stuff like this but I love piffers there definitely interesting and I think we can have a respectful convo.
Hello David I think your research is genuinely fascinating — I’m surprised more scholars aren’t taking it seriously.
That said, I do have some critiques. For instance, isn’t your model assuming that QST — or the quantitative trait loci involved — behave similarly across populations? Yes, 50,000 years is a relatively short time in evolutionary terms, but epistasis and how genes interact with each other can vary significantly. Your model seems solid when applied to European populations, because among Europeans we can actually demonstrate the additive contribution of alleles fairly well.
But the issue arises when we try to extend this to Africans or South Asians. I’m not convinced that the same alleles operate in the same way across these groups. Scaling a gene that affects one trait in one population might not work the same way in another, due to interaction effects. So while the QST vs. FST argument is clever, it still depends on the assumption that allelic effects are broadly conserved — which might not hold if divergence in background genome structure or epistatic interactions has occurred.
Another concern I have is about missing data — we lack genetic information from key ancient populations like those of Gandhāra, the Mauryan Empire, or the Mali Empire. Your research is heading in the right direction, and I respect that, but my critique is this: isn’t your model assuming that trait architectures interact in the same way across populations, and that group differences arise simply from the additive accumulation of alleles?
In reality, we might be dealing with broader similarities but divergent expressions — where the way a trait manifests in a population depends on more than just additive genetic effects.
Another issue I see that this study has is that it uses educational attainment which interacts with the environment per the scar Rowe idea I don't know if the Environments of heirtibility can be the same I think turkheimer warned of this.
But even if we are to grant that current the pgs only capture 5 to 15 percent which means the rest could be non casually linked to intellgence we might not know.
The biggest issue I see with using QST to infer selection is that it assumes additive expression of genetic variants — that alleles contribute to a trait in a linear, interchangeable way across populations. But QST is really just a statistical summary of between-group variance in predicted trait values. It doesn’t tell us how those variants are actually functioning in the biological context of different populations.
In other words, we’re using a shared statistical tool to infer shared biology, but the underlying biology might differ. The overall genome may be similar, but the specific loci, amino acid substitutions, and epigenetic regulation can vary enough to cause the same alleles — or even the same SNPs — to have different phenotypic effects across populations.
This is especially true for complex traits like cognition or educational attainment, where the trait isn't just molecular — it's embedded in development, neurobiology, and culture.
I understand you’ve pointed to robust cross-ancestry polygenic scores for schizophrenia or height to argue that effect sizes are conserved. But those are biologically simpler and more tightly defined traits. EA and IQ are much noisier, more culturally modulated, and more vulnerable to misattribution from indirect pathways.
So while I accept that QST shows divergence, I remain skeptical that this reflects selection on shared biology rather than selection — or drift — within population-specific architectures that just happen to produce correlated statistical scores
But TBF to you Kevin bird didn't have an argument he was mainly straw manning I think this an actual fair critique without straw manning you
To conclude
I think your studies makes a strong case, and it clearly shows a robust correlation — but until we explore more ancient DNA and improve current sampling from underrepresented populations, I remain agnostic. I’m not ruling your hypothesis out. I’m just saying we don’t yet know — and I think we should want to know. That’s what science should be about.
Btw In case your wondering I did use AI make the grammer easier to spread but core ideas are my own I just wanted to be clear the grammer correction is absolutely AI I have dyslexia and there is no way a sane person should have to read the horror of bad grammar.
Is there any genetic basis to the intellectual fecundity of classical Greece? Seems wild that Imperial Rome had peak scores because intellectually they are outshone by both Classical Greece and Modern nw Europe. Or would studies like this not pick up the cognitive elite production that produced the latter two?
IQ is not perfectly correlated to innovation (see the East Asian paradox). Personality and culture come into play as well. It seems like the Romans were a practical people so they used their IQ to optimise technologies that they borrowed from their neighbours and to build outstanding infrastructure for the time.
I'm interested in self-domestication, and think that the GFP may have been selected for, especially over time horizons of the last 15,000 years or so. Would love to see any analysis on that. Schizophrenia is another one that is heavily theorized to have to do with self-domestication
I have a paper under review analyzing a related topic, not GFP per se but the general factor of Arctic psychology. Will post about that later this month!
Looking forward to it! My thesis of gene-culture interaction intersects with ANE, as it so happens, as they are the originators of a type of culture that was useful for self-domestication, and spread worldwide. Joseph Campbell on the spread of a bullroarer mystery cult:
“Passing eastward across Siberia into America, as well as southeastwardly to Australia, shamanism traveled as but one element of a living compound that included—besides the x-ray style of animal painting and engraving, the atlatl, and the bullroarer—an elaborate complex of social regulations, ceremonials, and associated mythological ideas, which scholars have designated by the very broad term totemism.” ~Joseph Campbell, Historical Atlas of World Mythology, 1983
Am I misinterpreting or does this mean the average republic era Roman would have an IQ of ~120 if raised in a modern environment? Because that sounds absolutely insane, guess we now have yet another thing to envy Roma Aeterna for.
How might such current studies being biased toward European samples and relying on tag SNPs affect these scores?
Accuracy drops significantly outside the European cluster, so I wonder if this would also be affected by the greater variety of ancient samples.
I think as we move into more diverse whole genome datasets to run GWAS on (moving from <1M to ~3.2B bases), determining directly causal SNPs rather than only nearby SNPs affected by linkage disequilibrium, the accuracy and predictive ability of these scores will increase dramatically and remain higher across major genetic population clusters.
That's what I'm really looking forward to. But for now I'm curious how these scores might be affected, with potentially slightly less accuracy as we move further from modern samples due to the LD and tag SNP confounds.
There will be less accuracy, but it is not clear whether the error will be biased towards modern or older populations or if it will be random error, adding only noise.
The real question is why Britain_BA scored so high. Britain_BA is just a subset of Bell Beakers. Doesn't really make sense to attribute it to Britain_N, because these people were almost entirely eradicated by incoming continental Bell Beakers. Plus, Britain_BA is higher. Britain_BA was also one of the earliest highly lactase persistent cultures as well IIRC, it happened there way earlier than in the rest of Europe
The Yamnaya mostly conquered other Steppe Pastoralists such as the Khvalynsk and Srednii Stog. They made a few forays into the Balkans, but it was the Corded Ware Culture (~75% Steppe + 25% GAC) which expanded into Europe and formed the Bell Beaker Culture and the Indo-Iranic cultures. Would be very useful to have them, as well as the Globular Amphora Culture, Sintashta, Srubnaya, and Andronovo Cultures measured as well. Maybe also try the Sredni Stog samples in the Balkans from Nikitin et al.. I would imagine that the intelligence boom of the Iron Age was a product of selection under new, more hierarchical and exogamous Indo-European cultures similar to what happened with lactase persistence and hair/eye color
The Globular Amphora Culture is believed by some people to have invented the wheel, but it's debated. Early Indo-Iranians were pretty advanced, they invented the chariot and possibly also the composite bow. Peoples like them also probably played a pretty decent role in the composition of the Vedas. They were genetically basically Corded Ware-like, who were themselves genetically intermediate between Yamnaya and Bell Beaker. But, I don't think Bell Beakers would have been smarter due to admixture, as outside of Britain you seem to suggest Late Neolithic farmers weren't that smart. This could have been due to the resurgence in Hunter Gatherer ancestry.
Also, what are your thoughts on this paper questioning the reliability of PGS in ancient DNA? It just came out
I think a good way to measure would be to compare PGS for height and ICV with the actual estimated height and ICV of ancient skeletons. You would have to get access to the skeletons that each sample came from though, which is tough
the EA PGS for most of the groups you mentioned are already in the chart, so you can check where they stand. Regarding validation of the PGS using ancient skeletons, this has already been done: https://doi.org/10.1002/ajpa.24426
I also have a post on Height coming soon! Stay tuned!
I don't mean to be rude and I could be wrong, but this post feels written by AI. The wording and formatting reminds me almost exactly of how ChatGPT words things by default.
Do you have any ideas how many hours of serious work (data analysis, literature review, etc. ) goes into this? It is also my own original work, I don't even summarize other people's work! Be respectful
I respect your work immensely, I'm just pointing it out. I know that your work will reach more readers even if you still use AI if you can eliminate the AI-style phrasing. Personally, it is a very abrasive style.
Thank you. I chose a style that is synthetic and direct, because most readers do not have time to read lengthy posts. I will take your advice into consideration for the next posts
They were genetically very similar and there was gradual reciprocal cultural and genetic assimilation during the Republican period.The last three kings of Rome, from 616 to 509 BC, were of Etruscan origin and belonged to the Tarquin dynasty (Lucius Tarquinius Priscus, Servius Tullius, and Lucius Tarquinius Superbus)
Good post. It is refreshing to be moving away from the whole "weak Early European Farmer" narrative that has proliferated in misinformed online circles.
Our Neolithic European ancestors were patrilineal people. Patrilineality is a kinship system whereby an individual's family membership derives from the father's direct paternal lineage.
One of the earliest examples of patrilineality comes from Neolithic Europe in 4800 BC, where two families buried in a necropolis in Gurgy, France, are both connected via the male line as determined by the structure of the burial and Y-DNA haplogroup analysis.
Further evidence suggests that patrilineality was a kinship system used by Early European Farmers of indigenous European Hunter-Gatherer origin as well. The powerful elite buried within the Neolithic passage tomb at Newgrange belonged to Y-DNA haplogroup I2a (a Western Hunter-Gatherer lineage). You mentioned in the article that Early European Farmers in Britain underwent selection for intelligence.
Despite the Newgrange elite's overall autosomal ancestry being inherited from Anatolian Farmers and roughly a quarter from the Western Hunter-Gatherers, his phenotypic traits (specifically his complexion) resembled that of his hunter-gatherer ancestors several generations before him due to being the result of an incestuous relationship. Likely as a means by his parents to preserve the phenotypic traits of their hunter-gatherer ancestors.
The system of patrilineality for cultures primarily descended from European hunter-gatherers on the direct male line continued into the Middle Bronze Age.
Recent discoveries show that the use of patrilineality by farmer-descended populations was not affected by the introduction of Western Steppe Herder ancestry into the European gene pool. This is important to note because it does away with the misconstrued idea that the Early European Farmers simply vanished or whatever people say when the Western Steppe Herders arrived. A new paper by Allentoft et al. (2024) provides an example of an individual labelled NEO792 whose overall ancestry was 85% descended from Western Steppe Herders, but his Y-DNA lineage was I2a-S2703. A Funnelbeaker Early European Farmer subclade of Western Hunter-Gatherer origin.
NEO792 was buried in a megalithic tomb, not a burial mound, and carried mt-DNA haplogroup U2e2a1, a Western Steppe Herder derived female lineage. This means that despite NEO792 indirectly inheriting the majority of his overall ancestry from Western Steppe Herders, he was likely aware of his direct paternal heritage going back to his Funnelbeaker farmer male ancestors and was given a special megalithic tomb burial in the same fashion as them.
This shows us that patrilineality was a core social attitude of Bronze Age Europeans of Hunter-Gatherer origin and that their identity, cultural customs, culture, and overall perception of themselves was unique on a local level. Further evidence suggests that this is the case based on the aforementioned paper by Allentoft et al., where a new discovery shows us that the origin of the Germanic people lay with a core population of men belonging to Y-haplogroup I1, an indigenous European hunter-gatherer lineage, with additional hunter-gatherer ancestry.
The I1 cluster's introduction of stone cist burials into Denmark by 1750 BC shows us that a unique culture had developed within Scandinavia before or around 2600 BC and that this population would have likely used patrilineality as a means of kinship identification. This system and understanding of their ancestry would also be used as an exemplification of their elite lineage moving into the Nordic Bronze Age, where we see I1 become the leading lineage of the Nordic Bronze Age elite. Not a Western Steppe Herder lineage, but a hunter-gatherer lineage indigenous to Scandinavia itself.
I believe Kevin Bird was arguing that polygenic predictions work well for Europeans but are less effective for populations like Indians or Africans, who have distinct genetic profiles. East Asian genetics also seem to be severely underrepresented, at least based on what I've seen on Google.
He suggests that accurate predictions can be made for European populations, but these break down in non-European populations due to their differing genetic makeup.
Essentially we need to understand more about the theory of mind how G expands across cultures and if G is the final concept are there more intelligence?
Can we make vr tests to measure it's stuff like this but I love piffers there definitely interesting and I think we can have a respectful convo.
Thank you. Please check my latest post where I show that polygenic prediction in ancient DNA also works using East Asian GWAS
Hello David I think your research is genuinely fascinating — I’m surprised more scholars aren’t taking it seriously.
That said, I do have some critiques. For instance, isn’t your model assuming that QST — or the quantitative trait loci involved — behave similarly across populations? Yes, 50,000 years is a relatively short time in evolutionary terms, but epistasis and how genes interact with each other can vary significantly. Your model seems solid when applied to European populations, because among Europeans we can actually demonstrate the additive contribution of alleles fairly well.
But the issue arises when we try to extend this to Africans or South Asians. I’m not convinced that the same alleles operate in the same way across these groups. Scaling a gene that affects one trait in one population might not work the same way in another, due to interaction effects. So while the QST vs. FST argument is clever, it still depends on the assumption that allelic effects are broadly conserved — which might not hold if divergence in background genome structure or epistatic interactions has occurred.
Another concern I have is about missing data — we lack genetic information from key ancient populations like those of Gandhāra, the Mauryan Empire, or the Mali Empire. Your research is heading in the right direction, and I respect that, but my critique is this: isn’t your model assuming that trait architectures interact in the same way across populations, and that group differences arise simply from the additive accumulation of alleles?
In reality, we might be dealing with broader similarities but divergent expressions — where the way a trait manifests in a population depends on more than just additive genetic effects.
Another issue I see that this study has is that it uses educational attainment which interacts with the environment per the scar Rowe idea I don't know if the Environments of heirtibility can be the same I think turkheimer warned of this.
But even if we are to grant that current the pgs only capture 5 to 15 percent which means the rest could be non casually linked to intellgence we might not know.
The biggest issue I see with using QST to infer selection is that it assumes additive expression of genetic variants — that alleles contribute to a trait in a linear, interchangeable way across populations. But QST is really just a statistical summary of between-group variance in predicted trait values. It doesn’t tell us how those variants are actually functioning in the biological context of different populations.
In other words, we’re using a shared statistical tool to infer shared biology, but the underlying biology might differ. The overall genome may be similar, but the specific loci, amino acid substitutions, and epigenetic regulation can vary enough to cause the same alleles — or even the same SNPs — to have different phenotypic effects across populations.
This is especially true for complex traits like cognition or educational attainment, where the trait isn't just molecular — it's embedded in development, neurobiology, and culture.
I understand you’ve pointed to robust cross-ancestry polygenic scores for schizophrenia or height to argue that effect sizes are conserved. But those are biologically simpler and more tightly defined traits. EA and IQ are much noisier, more culturally modulated, and more vulnerable to misattribution from indirect pathways.
So while I accept that QST shows divergence, I remain skeptical that this reflects selection on shared biology rather than selection — or drift — within population-specific architectures that just happen to produce correlated statistical scores
But TBF to you Kevin bird didn't have an argument he was mainly straw manning I think this an actual fair critique without straw manning you
To conclude
I think your studies makes a strong case, and it clearly shows a robust correlation — but until we explore more ancient DNA and improve current sampling from underrepresented populations, I remain agnostic. I’m not ruling your hypothesis out. I’m just saying we don’t yet know — and I think we should want to know. That’s what science should be about.
Btw In case your wondering I did use AI make the grammer easier to spread but core ideas are my own I just wanted to be clear the grammer correction is absolutely AI I have dyslexia and there is no way a sane person should have to read the horror of bad grammar.
Is there any genetic basis to the intellectual fecundity of classical Greece? Seems wild that Imperial Rome had peak scores because intellectually they are outshone by both Classical Greece and Modern nw Europe. Or would studies like this not pick up the cognitive elite production that produced the latter two?
IQ is not perfectly correlated to innovation (see the East Asian paradox). Personality and culture come into play as well. It seems like the Romans were a practical people so they used their IQ to optimise technologies that they borrowed from their neighbours and to build outstanding infrastructure for the time.
I'm interested in self-domestication, and think that the GFP may have been selected for, especially over time horizons of the last 15,000 years or so. Would love to see any analysis on that. Schizophrenia is another one that is heavily theorized to have to do with self-domestication
I have a paper under review analyzing a related topic, not GFP per se but the general factor of Arctic psychology. Will post about that later this month!
Looking forward to it! My thesis of gene-culture interaction intersects with ANE, as it so happens, as they are the originators of a type of culture that was useful for self-domestication, and spread worldwide. Joseph Campbell on the spread of a bullroarer mystery cult:
“Passing eastward across Siberia into America, as well as southeastwardly to Australia, shamanism traveled as but one element of a living compound that included—besides the x-ray style of animal painting and engraving, the atlatl, and the bullroarer—an elaborate complex of social regulations, ceremonials, and associated mythological ideas, which scholars have designated by the very broad term totemism.” ~Joseph Campbell, Historical Atlas of World Mythology, 1983
REEEEEE replicate this paper with the tan et al FGWAS weights
https://openpsych.net/paper/59/
Am I misinterpreting or does this mean the average republic era Roman would have an IQ of ~120 if raised in a modern environment? Because that sounds absolutely insane, guess we now have yet another thing to envy Roma Aeterna for.
More like 110 or so but with a lot of IFs
How might such current studies being biased toward European samples and relying on tag SNPs affect these scores?
Accuracy drops significantly outside the European cluster, so I wonder if this would also be affected by the greater variety of ancient samples.
I think as we move into more diverse whole genome datasets to run GWAS on (moving from <1M to ~3.2B bases), determining directly causal SNPs rather than only nearby SNPs affected by linkage disequilibrium, the accuracy and predictive ability of these scores will increase dramatically and remain higher across major genetic population clusters.
That's what I'm really looking forward to. But for now I'm curious how these scores might be affected, with potentially slightly less accuracy as we move further from modern samples due to the LD and tag SNP confounds.
There will be less accuracy, but it is not clear whether the error will be biased towards modern or older populations or if it will be random error, adding only noise.
The real question is why Britain_BA scored so high. Britain_BA is just a subset of Bell Beakers. Doesn't really make sense to attribute it to Britain_N, because these people were almost entirely eradicated by incoming continental Bell Beakers. Plus, Britain_BA is higher. Britain_BA was also one of the earliest highly lactase persistent cultures as well IIRC, it happened there way earlier than in the rest of Europe
The Yamnaya mostly conquered other Steppe Pastoralists such as the Khvalynsk and Srednii Stog. They made a few forays into the Balkans, but it was the Corded Ware Culture (~75% Steppe + 25% GAC) which expanded into Europe and formed the Bell Beaker Culture and the Indo-Iranic cultures. Would be very useful to have them, as well as the Globular Amphora Culture, Sintashta, Srubnaya, and Andronovo Cultures measured as well. Maybe also try the Sredni Stog samples in the Balkans from Nikitin et al.. I would imagine that the intelligence boom of the Iron Age was a product of selection under new, more hierarchical and exogamous Indo-European cultures similar to what happened with lactase persistence and hair/eye color
The Globular Amphora Culture is believed by some people to have invented the wheel, but it's debated. Early Indo-Iranians were pretty advanced, they invented the chariot and possibly also the composite bow. Peoples like them also probably played a pretty decent role in the composition of the Vedas. They were genetically basically Corded Ware-like, who were themselves genetically intermediate between Yamnaya and Bell Beaker. But, I don't think Bell Beakers would have been smarter due to admixture, as outside of Britain you seem to suggest Late Neolithic farmers weren't that smart. This could have been due to the resurgence in Hunter Gatherer ancestry.
Also, what are your thoughts on this paper questioning the reliability of PGS in ancient DNA? It just came out
https://www.sciencedirect.com/science/article/abs/pii/S0002929725001909
I think a good way to measure would be to compare PGS for height and ICV with the actual estimated height and ICV of ancient skeletons. You would have to get access to the skeletons that each sample came from though, which is tough
the EA PGS for most of the groups you mentioned are already in the chart, so you can check where they stand. Regarding validation of the PGS using ancient skeletons, this has already been done: https://doi.org/10.1002/ajpa.24426
I also have a post on Height coming soon! Stay tuned!
I don't mean to be rude and I could be wrong, but this post feels written by AI. The wording and formatting reminds me almost exactly of how ChatGPT words things by default.
Do you have any ideas how many hours of serious work (data analysis, literature review, etc. ) goes into this? It is also my own original work, I don't even summarize other people's work! Be respectful
I respect your work immensely, I'm just pointing it out. I know that your work will reach more readers even if you still use AI if you can eliminate the AI-style phrasing. Personally, it is a very abrasive style.
Thank you. I chose a style that is synthetic and direct, because most readers do not have time to read lengthy posts. I will take your advice into consideration for the next posts
I thought the Etruscans were a distinct people from the latter Romans?
They were genetically very similar and there was gradual reciprocal cultural and genetic assimilation during the Republican period.The last three kings of Rome, from 616 to 509 BC, were of Etruscan origin and belonged to the Tarquin dynasty (Lucius Tarquinius Priscus, Servius Tullius, and Lucius Tarquinius Superbus)