Almost a decade ago scientists at the David Reich Lab extracted DNA from the remains of three men from the Khvalynsk II cemetery at the northern end of the Pontic-Caspian (PC) steppe.
These Eneolithic Eastern Europeans showed significant genetic heterogeneity, with highly variable levels of Eastern Hunter-Gatherer (EHG) and Near Eastern-related ancestry components.
As a result, the people at the David Reich Lab concluded that the Eneolithic populations of the PC steppe formed from a relatively recent admixture between local hunter-gatherers and Near Eastern migrants.
Unfortunately, this view has since become the consensus among scientists working with ancient DNA.
I say unfortunately because there's a more straightforward and indeed obvious explanation for the genetic heterogeneity among the samples from Khvalynsk II. It's also the only correct explanation, and it doesn't involve any recent gene flow from the Near East.
Here it is, in point form, as simply as I can put it:
- EHG is best represented by samples from Karelia and Lebyazhinka, which are modern-day Russian localities in the forest zone and on the border between the steppe and the forest-steppe, respectively
- Khvalynsk II is also located on the boundary between the steppe and the forest-steppe, and very far from the Near East
- so the genetic structure of the people buried at Khvalynsk II does represent an admixture event
- however, this admixture event simply involved an EHG population from the forest-steppe and a very distantly Near Eastern-related group native to the steppe (that is, two different Eastern European populations).
I've written this blog post because I think David Reich, Nick Patterson, Iosif Lazaridis and colleagues should finally admit that they didn't quite get this right. And it'd be nice if they could put out a paper sometime soon in which they set the record straight.
See also...
450 comments:
«Oldest ‹Older 401 – 450 of 450@Samuel Andrews,
Do you have an email address?....I have a few questions about mtDNA.
sammyisaac107@gmail.com
@J.S. sure, here are the samples from the latest G25 files I have: https://pastebin.com/941M6tLr
I've relabelled some of them slightly, just so they all follow the same naming conventions with each other.
@J.S. also may be of interest; quick survey of closest population matches for all those France IA->RomanP->MA samples - https://imgur.com/a/AxmOLGF
@Livonia, could you give some more breakdown of who it would be useful to have as an extra Celtic population on the PCA (e.g. England_IA, present day populations ?)
@Matt
Thank you very much! Can you have a look at this one?
FRA_Brittany_Morbihan_EBA:I15033,0.124067,0.139128,0.059585,0.041021,0.051702,0.01004,0.003995,0.004384,0.020452,0.022051,-0.00065,0.007943,-0.01115,-0.022157,0.012893,0.018828,0.01017,-0.000633,0.003645,0.010755,0.004866,0.003957,-0.008874,-0.005663,-0.000239
@Livonia, forget my last question, realized what you mean. 'Celtic' in Vahaduo is present day samples from Welsh and Irish.
So here is what I think you would be looking for: https://imgur.com/a/GspN5N6
@J.S. Here: https://imgur.com/a/peB4efc
Seems to have relatively high similarity to EBA people from NE France, SW Germany and later people from France (e.g. ESP_Hellenistic_oLocal 300 BCE = Gauls in the Hellenistic colony).
@Ramber
https://drive.google.com/file/d/1RiqKlF8n5XeuK_XGV-IzeeDBe_lF9OSy/view?usp=sharing
@Vara
The Kunga were used way before the domesticated horse reached the Near East, and were first mentioned c. 2600 BCE.
@Davidski
Is this from the second genotype data? Thank you so much! Really appreciated it!
@Michalis Moriopoulos
I know that you are busy due to your upcoming Valentine's Day release, but can you also help annotate the second spreadsheet Davidski just sent to me?
@Matt
Thanks a lot for you help. Still awaiting for Fanny Mendisco et al. paper "Genomic ancestry during Antiquity in France".
@ Matt
Wonder if that population from Southwest Germany/Northeastern France was distantly related to the Amesbury Archer ?
@Davidski,
A top priority I think should be to convert the 23,000 year old Solutrean.
Not random Mongolian DNA samples :)
@Rob The qpAdm (table s12) shows 20,9% Magdalenian-like ancestry and 42,5% EHG-like ancestry with a pValue of 0,66
I have no reason to assume this is not the case. Perhaps submit a response article if you think the calculation is incorrect
@Samuel Andrews,
Did you get my email?
@Rob I'm sure that sounded a lot smarter when you were reciting it in front of your bathroom mirror
@ Opheus
“I have no reason to assume this is not the case. Perhaps submit a response article if you think the calculation is incorrect”
Yes it’s clear you don’t see things very well, and no I don’t have to “submit a letter” or go on Oprah
I will simply present the facts
Rou_MES BEL_GoyetQ2 ITA_Tagliente_Late_Paleo RUS_Sidelkino_HG
0.0 52.2 47.8
There is no Magdalenian in either this sample, nor any other in Romania.
The Magdalenian did not make it beyond the Carpathians, so if they claimed so it would be historically inacurate and mean something is probably wrong with their claim. In this case, the authors made a simple error - they coloured the pie-chart wrong. It should be WHG + EHG, just like every other Serbia/ Romanian_HG, not WHG + GoyetQ2
The thesis of Luka Papac (the author who gave us the very detailed Czech Early Corded Ware transect and more) has been uploaded by Planck:
https://pure.mpg.de/pubman/faces/ViewItemOverviewPage.jsp?itemId=item_3489057 – “Tracking population history, social structure and intergroup exchange in Neolithic to Bronze Age Europe using ancient human and virus genomes”
phd thesis for free download – https://www.db-thueringen.de/receive/dbt_mods_00053720
“Abstract: Archaeogenetics has revealed two major migrations that shaped European prehistory; one starting in the 7th millennium BCE and associated with the transition to agriculture, and another with the appearence of the Corded Ware cultural complex in the early 3rd millennium BCE. However, most findings have come from studies employing a continent-wide and/or low regional density sampling approach, leaving many temporal and geographic gaps in the current European archaeogenetic record. In order to build upon previous studies, this thesis employed a high-density sampling approach to study the population structure of Neolithic to Bronze Age central Europe in finer detail.
In total, genome-wide ancient DNA was analysed from 283 newly reported individuals from present-day Bohemia, Czech Republic. In doing so, new insights into population structure, genetic turnovers, and social processes were discovered and discussed. These include a genetic turnover between the Jordanow and Funnelbeaker cultural groups, a genetic turnover between Early and Late Corded Ware groups, and a genetic turnover between Bell Beaker and Únětice groups. Insights into social structure and processes revealed the likely contacts between early farming communities with hunter-gatherer groups of Eastern Europe, that the integration of pre-Corded Ware people into Corded Ware society was likely a sex-biased process, and the emergence of a strict patrilineal social system in Late Corded Ware and Bell Beaker cultural groups which broke down with the appearance of early Únětice cultural groups.
Insights from a detailed investigation into an Early Bronze Age cemetery from Mikulovice revealed a genetically homogeneous group with evidence of patrilocality and families spanning up to four generations. This thesis shows the potential of more detailed, higher-resolution studies to reveal previously unknown genetic structure and social processes.”
May be some more detail than we saw in the shorter paper.
@All
I'm working on a post that definitely explains what Yamnaya is.
It should be ready in a couple of days, and will lay to rest all these bullshit claims about its Armenian or Central Asian origins.
@Davidski
Actually, you should show your f3 output to Nick Patterson and tell him that you found evidence of the Central Asian origins of Yamnaya.
If he doesn't die from laughing too much, he might explain to you that you're misinterpreting the output.
Basically, what's happening is that f3 stats produce more extreme results when they have more divergent populations to work with. So unfortunately, they're useless if you want to look at complex and fine scale models
There's a technical explanation for this glitch. Would you like to hear it from me or someone at the Reich Lab?
f3-statistics can give negative results when the populations taken as sources are very different from each other. Also, the f3 statistic can be positive when the share of one or both components is not high, for example, they are mixed in the proportion of 90% + 10%, or 95%+5%, or 20%+20% ((that is, when there are more than two components). But it cannot be positive when the proportion between the components is approximately 50/50 or 60/40, as in the case of the EHG-like + CHG-like components, as modeled for Yamnaya culture. If you have specific described cases where, when mixing components in a 50/50 ratio, f3-statistics gives a false positive result, I will be glad to hear them. In addition, if you don't like f3-statistics, you can look at the results of qpadm, taking the Progress component, Neolithic of Turkmenistan or Sarazm as the CHG-source for Yamnaya culture.
I don't quite understand your version - that is, you think that the CHG-like admixture in the Yamnaya culture evolved from the local Paleolithic of the Caucasus, and did not come from Central Asia? Can you somehow trace it back to specific archaeological cultures?
@Tarbagan
You just claimed that Yamnaya was derived from Central Asia and Progress from Shulaveri.
This tells me that you're a total moron.
You should stay away from f3 statistics or any other statistics and find yourself a much less complicated hobby.
The Wikipedia entries about Yamnaya and Western Steppe Herders are utter garbage.
@Matt Interesting, early CWC had a higher ydna diversity signifying no strict patrilinieality and/or social hierarchy, something Zegarac et al 2021 noticed as well for a BA Serbian culture. Wonder if the R1a spread is correlated with the tribes who mixed with EEFs first, possibly leading to acquiring superior technology and using that to dominate the rest of CWC. A shift to a sedentary farmer culture qould point at that too, with farming>more resources>more power and offspring (confirmed by Fibiger). The GAC study also comes to mind where all the males belonged to a single lineage and thus strict patrilineality. Chintalapati et al 2022 notes low exogamy in CWC (per quote, very little further mixture) so this happened early on. Unetice being more "egalitarian" and originating from the Northeast also seems to support this hypothesis.
Almost the opposite of what Gimbutas guessed lol (especially in the light of the Fibiger study which will be adressed soon), but Lazaridis & Reich 2017 already adressed this for central Europe BA where they found no mating bias in admixture.
@Davidski
"@All
I'm working on a post that definitely explains what Yamnaya is.
It should be ready in a couple of days, and will lay to rest all these bullshit claims about its Armenian or Central Asian origins."
Thanks! I am looking forward to it.
Could you possibly throw in your opinion on the origin of Corded Ware vis-à-vis its apparent genetic relationship to Yamnaya? I know you have already done that here and there in your writings, but it would be really nice to have a compact summary in the same post in which you explain the genetic origin of Yamnaya.
@Davidski
Where did I say that Yamnaya culture came from Central Asia? The Yamnaya culture originated in Eastern Europe, most likely between the lower reaches of the Don and the Volga. I was talking about the fact that the CHG-like component of the Yamnaya culture most likely has a Central Asian, not a Caucasian, origin. And it came a long time ago, somewhere in the Paleolithic or Mesolithic. And this same component formed a CHG-like admixture in the Khvalyn culture. Anthropologists identify a very close proximity of some skulls of the Yamnaya and Khvalynsk cultures to the craniological series of Turkmenistan. In Eastern Europe, this CHG-like component was mixed with EHG and with a cluster of Ukrainian hunters, which formed the genetics of Yamnaya and Khvalyn cultures. And I never understood that you are not satisfied with this model.
@Andrzejewski
"The Wikipedia entries about Yamnaya and Western Steppe Herders are utter garbage."
Well then re-write it. A lot of crap remains on wikipedia for years just because no one can be bothered to change it.
@Orpheus, I'm not sure about advanced technology from EEF, but yeah, there must be some reason for the reducing diversity. It seems plausible to me that some combination of farming to boost population size, and mobility or use of pasture from steppe groups was superior to a purer pastoral lifestyle in Europe and would win out. You're probably right to talk about setting patrilineality in wider context of the same practice in EEF groups (e.g. esp megaliths, GAC), Yamnaya, etc.
I'm not sure about the point about mating admixture; it seems to me like this was present for early Corded Ware culture, and so becomes visible in the y-chromosome replacements, but the picture may be less clear as we get to areas of Southern Europe, and also the linkages between autosomal steppe ancestry and y-dna may break down in ways that mean y-dna shifts are not consistently associated with the same X chromsome:autosome ancestry ratio shifts.
...
One other thing to note about this thesis is that Papac's analysis of ancient HBV dna reveals who TTK002 was:
“Archaeological background: Here we provide information on the archaeological context for some of the samples analyzed in this study (for the rest of the samples, see references in Data S1).
Tutkaul”, Tajikistan. (Individuals TTK002): The site of Tutkaul is located in southern Tajikistan, 70 km southeastwards of Dushanbe in the Dashti-Mazar region. The site was discovered during the archaeological survey of the Nurek dam’s flooding area led by A.P. Okladnikov in 1956 (106). Excavations were conducted during six field seasons (1963,1965-1969) as part of a rescue archeological program led by the Tajik Archeological Expedition (107). The upper part of the stratigraphy consists of a medieval fortified settlement, followed by levels 1 and 2 which are attributed to the Hissar Neolithic culture.
The lowest stratigraphic units (3 and 2a) belong to the Early and Late Epipaleolithic (108). At the base of level 2, three burials were identified containing the remains of four individuals: a female adult (burial nb. 1), a subadult (burial nb. 2) and two children (burial nb. 3). The burials were oriented to the SE-NW, the skeletons were in a bent condition on the left side, suggesting that the bodies were tied up before being buried. TTK002 is the vertebra from a child found in burial nb. 3. An unidentified bone fragment from the same burial was radiocarbon-dated to 8425-8025 cal BP (GV-02104 7450±106)."
TTK002 = Tajikistan Late Epipaleolithic 6225 BCE.
There are also a couple of things noted about the samples from Yu 2023 (still unpublished), KVH and JAZ:
"Khvalynsk, Volga region, Russia (KVK001): KVK001 (museum nb. KO128) is a human skull fragment that was discovered in 1927 on the Khoroshensky Island of the Volga River near the town of Khvalynsk, by the local museum staff. It was found in a semisubmerged pebble layer, which also contained abundant remains of Quaternary fauna (101). Weinert (102), who first published the finding, attributed it to a human from the Paleolithic.
Later, Gremyatsky (103) concluded that it belonged to a, possibly male, anatomically modern human with some Neanderthal features. As noted by Bader (104), the stratigraphy of alluvial deposits in the region is quite uniform and stable. After conducting additional analyzes, the researcher did not rule out that this skull fragment could be associated with the Upper Paleolithic (Aurignac-Solutré). While the pebble layer in which it was found, was attributed to the Mindel-Riss Interglacial, A. P. Ososkov (105) considered it possible that the sandstone load of this region was formed during the Late Paleolithic, and the results of radiocarbon dating indicated that the sample belongs to the Early Mesolithic time (OxA-23001, 9045 ±40 uncal BP)."
Uncal BP, but provisionally this sample could be called RUS_Khvalynsk_Meso, 7050 BCE.
"Jazikovo (Yazykovo), Volga region, Russia (JAZ001): The remains of an adult woman (JAZ001, museum nb. 8619) were excavated at the Yazykovo peatland near the village of Yazykovo, Kashinsky district, Russia, in a cultural layer which was firstly discovered by B.S. Zhukov, A.E. Alikhova and M.V. Voevodsky in 1928 on the banks of the Yakhroma river. The researchers discovered ceramics, the stratigraphic position of which indicated that they were anterior to the Pit-Comb Ware culture (124). This pointed to Yazykovo being one of the most ancient Neolithic sites in the Upper Volga region.
Today, Yazikovo encompasses three complexes (Yazikovo 1,2 and 3), which are attributed to the Central variant of the Upper-Volga archaeological culture (125). Based on the analysis of ceramics from different sites including Lyalovo, Yazykovo 1 and 2, Nikolo-Perevoz 1 and 2, Kholomonikha, Lipki, Volosovo 1, Malo-Okulovskaya 1, Malo-Borskaya and Balakhninskaya 3, Zhukov identified two variants characterizing the development of Neolithic societies in the Upper Volga region (124). The variant presumed to be the oldest one was defined as the “Yazykovo” type: thick-walled pottery with abundant grit in the dough, decorated with parallel carved stripes. The second variant was denominated as the “circular-pit comb complex”, due to similarities with the “comb” pottery and particular ornamental features. Further studies have led to a more detailed periodization of the Volga- Oka Neolithic, which divides it into an “early” and “developed” period. In the Upper Volga region, subsistence strategies were maintained across the Mesolithic to Neolithic transition, and the beginning of the Neolithic is solely marked by the appearance of pottery.
The Early Neolithic period extends from 7100/7000 to 6100/6000 uncal BP (126), and is viewed as a period in which Yazykovo-type ceramics were imported in the region through occasional trade with non-autochthonous individuals, while the local lithic industry remained unchanged (127–129). The developed Neolithic period extends from 6100/6000 to 3800/3700 uncal BP (130), and is marked by the appearance of, likely locally produced, comb-ornamented ceramics, possibly reflecting population influx.
JAZ001 was radiocarbon dated to 6329 ±21 uncal BP (MAMS-37910), thus placing it at the end of the Early Neolithic period."
RUS_Upper_Volga_Jazikovo_Neolithic, 4330 BCE?
On PCA and closest neighbour via G25: https://imgur.com/a/yikVoNM
Note that the map of samples for ancient HBV genomes and data from Papac's thesis on HBV also confirms that DOG007 is from Doggerbank at 7500 BCE (ascribed to the Netherlands as a country, but in the North Sea on the map) and that the samples labelled MN2003 etc are from Minino at 8600 BCE.
@Andrze
The mainstream hasn't yet acknowledged that R1b1 and R1a are European lineages.
23andme tells its customers that R1b comes from the Fertile Crecent.
No one has sounded the alarms that HEY WE HAVE LOTS OF R1b1 and R1a from Mesolithic Europe. HEY!!!
@Andrze
Yeah, it's annoying. Wikipedia, everyone the mainstream, is so outdated.
@Andrze,
You should Follow me on Telegram. I have 1k followers.
https://t.me/genoshistoria
I talk about stuff like that all the time.
This idea that Yamnaya has "substantial" Armenian-related ancestry is totally wrong. It's about as stupid as claiming that R1a is from India.
I'll explain why in my new blog post.
We need to think about some effective ways to make Iosif Lazaridis, David Reich and Nick Patterson admit that they're wrong about this, and get them to produce a correction.
@Matt Farming technology is the first thing coming to mind, after all CWC were farmers (Lamnidis et al 2018, Saag et al 2021). An early CWC group amassing resources and expanding is not far fetched at all.
Additional technology could come from trading (especially after they increased their resources) with nearby EEF cultures which would lead to more carts and livestock, and possibly more weapons (stone axes, adzes and arrowheads, flint knives, wooden and stone-headed clubs, antler picks, sling shots were all found in Neolithic farmer cultures as per Fibiger). These conditions also lead to reduced infant mortality compared to non-farmer pastoralists. So a population growth followed by outcompeting the other CWC groups, possibly with raids or/and warfare involved.
If the EEF admixture in CWC appears early on but stabilizes quickly and does not increase, as Chintalapati et al 2022 noted, this could serve as additional evidence. But nothing concrete obviously
@Davidski Considering they aren't going to read your blog, you have two options to draw their attention: email them (will probably get ignored) or gather all the arguments and sources you have and publish a paper or two if you can, probably in response form (see Lazaridis & Reich 2017 for example) since it's easier than a lengthy study.
What else is there to do which will force an answer? As much as I like the drama, I doubt the 150th dear Iosif post is gonna be the charm
@ Matt
“ same practice in EEF groups (e.g. esp megaliths, GAC),”
These groups should not be referred to as “EEF”- that should be reserved for Balkan N, LBK, Italian Impressa
GAC had a more mobile, cattle oriented economy , with carts / wagons etc
@Davidski,
I can't think of way to say it without offending them.
Is it even true that Yamnaya and the like are 50% CHG? Aren't they closer to 30%?
“GAC had a more mobile, cattle oriented economy , with carts / wagons etc”
Also additional 20-30% HG ancestry, different Patriarchs, etc
A bit of a Eureka! moment here.
I worked out how to show what Yamnaya/Corded Ware is with no ambiguity. And it's definitely not Armenian-related.
Blog post tomorrow. Oh dear Iosif!
@Davidksi
"A bit of a Eureka! moment here.
I worked out how to show what Yamnaya/Corded Ware is with no ambiguity. And it's definitely not Armenian-related.
Blog post tomorrow. Oh dear Iosif!"
Cool!
Glad to hear you're including Corded Ware.
Looking forward to tomorrow's blog post. Thanks!
Hell yeah !!!
@David
The issue with your EEF theory of J2B2-L283 is that we don't see its parents anywhere in Neolithic Europe, Anatolia or the Middle East. We've had hundreds of EEF/ANF samples and there isn't a single J2B2-L283 in sight, nor a predecessor. With E-V13 for example, even though it's rare, we see its E-L618 parent in ancient Greece, Hungary, Croatia, Spain, etc...
J2B2-L83 just pops out of anywhere in Europe during the Bronze Age, and it's always associated with higher steppe ancestry. This points to a CHG-related Y-DNA that was a minor steppe lineage.
I think it's pretty clear that so far the steppes are poorly sampled compared to Neolithic Europe. J2B2 aside, we still need to find all the related L51 and R1a hotspots that we know for a fact exist somewhere on the steppe. This tendency to just think of Yamnaya only as Z2103 carriers should be abandoned.
Wishful thinking can only be combated with scientific data. Of course there are L283 with zero steppe ancestry (Bronze Age Sardinia, Kabardino-Balkaria and Armenia). L283 in the Mycenaean Bronze Age and V13 in the Thracian Iron Age only means that those lineages were part of IE speaking human groups, not that those lineages have steppe origin (they do not exist in the European genetic record until the Bronze or Iron Age). It is the same case as L51 and M417, there are still people who think that they were scarce or low social class markers and that is why they have not been found in the steppe kurgans, the reality is much easier to explain, neither L283 nor V13 nor L51>L151 nor M417 participated in the Yamnaya culture, and whoever says the opposite will have to prove it.
Some stubborn mathematicians try to give lessons in genetics, archeology and linguistics to everyone, the longer they take to rectify the harder the fall will be.
@DragonHermit
"we still need to find all the related L51 and R1a hotspots that we know for a fact exist somewhere on the steppe. This tendency to just think of Yamnaya only as Z2103 carriers should be abandoned."
Ever heard of Corded Ware? Yamnaya isn't ancestral to all IE off shoot cultures so there doesn't have to be non R1b-Z2103. There is more than just Yamnaya in the Pontic-Caspian steppe, we have an extensive overview of the Yamnayan uniparentals and to say that R1b-Z2103 is the dominant Yamnayan haplogroup is an archaeogenetic fact. What we don't have (yet) is a basic overview of Suvorovo-Novodanilovka, Cernavoda, Usatovo etc. namely archeological complexes which predate Yamnaya.
@Gaska
All those J2B2-L283s you mentioned are downstream branches. The TMRCA so far is 3500-3600 BC. Based on that logic, we've also found R1b in the Middle East with almost no steppe. It means nothing.
@Mark
Ever hear that CW is modelled as Yamnaya + EEF and IBD shows they were part of the same population? They were literal cousins speaking the same languages.
The fact that linguists still can't distinguish between "Yamnaya languages" like Greek, Armenian and Albanian, and "CW languages" like Italo-Celtic, Balto-Slavic, Germanic, shows that there really is no difference between Yamnaya and CW ancestors. Any significant bifurcation, and it would show up linguistically. In fact, the latest IE linguistic megastudy pinned Italo-Celtic as the most archaic/exotic branch, meaning Germanic/Balto-Slavic are closer to Greek, Armenian and Albanian.
@Livonia I’m having trouble understanding your post about Gaulish and Insular Celtic “crossing in 1300 CE”.
Is that date a typo?
New ancient IBD tool out. This will be MASSIVE for the genetics community. I wish I had some free time this year to delve into this stuff, but hopefully @Davidski does and incorporates this into his analysis. Within a few hundred years time frame we can detect gene flow, including potential geneflow between Yamnaya and Caucauses/Anatolia (sample quality and timeframe permitting).
For the paper, the main takeaway is the obvious (D. Reich already talked about this):
"We find that the first individuals in Central and Northern Europe carrying high
amounts of Steppe-ancestry, associated with the Corded Ware culture, share high rates of long IBD (12-25 cM) with Yamnaya herders of the Pontic-Caspian steppe, signaling a strong bottleneck and a recent biological connection on the order of only few hundred years, providing evidence that the Yamnaya themselves are a main source of Steppe ancestry in Corded Ware people"
There are still people in other forums arguing about "Srdeny Stog PIE" who don't understand how IBD works LOL. This is a 23andme-like tool. These tools are used to discover close cousins, not 1000 to 2000 years apart. Yamnaya and CW were very closely related in their days, not going back to pre-horse domestication days.
Main text: "Karelia and Lebyazhinka, ... in the forest zone and on the border between the steppe and the forest-steppe, respectively"
<Sorry, both are by far deep north in the forrest zone.
archlingo
Lebyazhinka is on the northern border between the forest steppe and steppe.
It's near Samara where many Yamnaya samples come from.
Post a Comment