The history of this haplotype is as follows. Note that the numbers that follow are all approximate, and I have only mentioned what is considered the defining mutations in each haplogroup. Many other mutations are equivalent to it, i.e., they happened at roughly the same time as far as we know today. Also, I haven't gone back to the sources to find the error bars. Also, here I present only the lineage that leads to me, a summary of other lineages which populated India can be found in my pages dealing with the history of Bengal. In a similar vein, this description only traces back to at most a 100,000 years back, you can find a cursory account of human evolution dealing with the period before that here instead. The data that is not from reading the original papers comes from the summary at the isogg web site.
As the Y chromosome is passed on, small ‘mutations’ or copying errors (given names like M17) occur, dividing all of us into haplogroups (e.g., R1a) who share male ancestry. From the population living 60–90,000 years back, all lineages have died out but one, so the common ancestor of the remaining lineage is said to be "Y-chromosome Adam". By 60,000 years, the A haplotype was well established. Among these people, the people carrying the mutation L1085 were defined as the haplogroup A0-T from whom all groups other than A00 arose. The A1 haplotype was then defined as the descendents carrying the mutation P305 that may have arisen aroung 140,000 years back, and its subgroup A1b by P108, which may have arisen around 110,000 years back among these. Around 55,000 years back (to within large errors), my ancestors got the mutation M91 somewhere in north-east Africa and became the BT haplotype. With a further mutation M168/PF1416 (a / separates alternate names for the same marker; a .1, .2, etc. marks various times the same locus mutated), they distinguished themselves from the haplogroup B, and became the haplogroup CT. Next the mutation P143/PF2587 arose somewhere in northeast Africa around 31–55,000 years ago and made us into the haplogroup CFxDE, whose descendants were the primary homo sapiens to populate the world outside Africa. The founder of this lineage is, therefore, often fondly called the Eurasian Adam. All we know so far is he lived some 2500 generations, give or take a 1000, before me; and his descendants had a wanderlust, possibly nomadically following the animals during a period of moist climate, or possibly driven by a population increase resulting from the favourable climate. Whatever be the reason, we see an increasing sophistication in tools and appearance of art as the generations progressed.
This branch of the population that came out of Africa started moving eastward along the coast, whereas the other group called DE and defined by M145/P205/PF144 that arose about 50,000 years back, took a more northerly route. The descendants of both of these are present today in central and southeast Asia (E went mainly to Europe and D has been displaced from most of India except for Tibet and Andaman—both of which are very high in D), and descendants of both these groups have migrated back to Africa at various times. After a wave of emigrants (C; defined by M130/Page51/RPS4Y711) left around 50000 years back to ultimately populate the southern coast of Asia all the way to Australia (and also Central Asia and Americas), the marker M89 made its appearance and we started being called the haplogroup F. Most non-Africans are actually the descendant of an individual carrying these mutations (about 2500 generations back, possibly 60–80 thousand years ago if it arose in Africa), who lived somewhere in northern Africa or just outside near the middle east. This group probably had a rapid expansion around 41–52,000 years back. Some of his descendants (G; marked by M201/PF2957 that probably arose around 10,000–23,000 years ago, but could be as early as 76,000 years back) may have lived in the northern part of the middle east. The rest are called HIJK and defined by F929/M578/PF3494/S6378. Most of the descendents went through Iran into either India (H; marked by L901/M2939) or the Steppes of Central Asia (IJK), possibly hunting game like buffalo, antelope, and wooly mammoths. These IJK were marked by L15/M523/PF3492/S137, which arose around 45000 years back, probably while still in the middle east, or in Africa, because some of his descendants (IJ; marked by M429/P125/PF3535 that arose around the same time) migrated to both Europe (I) and Asia (J).
Around 40,000 years back, and say 1500 generations back, IJK people living in Southwest Asia, may be Iran, developed the markers M9 and all of us who descended from him are classified as haplogroup K. His descendants are popularly called the Eurasian clan, and are a vast majority in Asia, Europe, and the Americas, though some ancient K are also found in Africa.
A further mutation M526 separated the group called K2 or KxLT, whose descendants went to Central Asia, from the group called LT=K1 marked by L298/P326 whose descendents were the middle eastern T (marked by M184/Page34/USP9Y+3178) and the Indian L (marked by M20/PF5570). The next major marker is M1221/P331/PF5911 made us into K2b, to be distinguished from the NO=K2a clade marked by M214/Page39 that probably arose around Mongolia. Then, P295/PF5866/S8 that arose among the K2b people marked us as belonging to haplogroup P; as opposed to K2b1 marked bu P399 whose descendents are the east asian M (marked by P256) and S (marked by B254). The mutations defining P probably arose somewhere in central Asia, may be as far South as the Pamir region, or north as Uzbekistan, Kazakhstan or even Southern Siberia, about 35000 BC, still more than a 1000 generations before us. Their descendants started in these game rich regions north of the Hindukush, but ultimately populated most of Europe and the Americas; even while surviving the ice age that was now upon them.
The appearance of M45/PF5962 made us P1. Then M207/Page37/UTY2 somewhere in Asia around 27000 years ago then made us into group R (also called R-M207) and distinguished us from Q (marked by M242, whose descendants went through the Altai/Baikal region), and M173/P241/Page29 around 18500 years back in southwestern Asia made it R1 (also called R-M173) (R2, the other defined haplogroup, is marked by M479/PF6107—this is the new nomenclature from 2010). The descendants of these people brought cave paintings to Europe. Reality, of course, could be more complicated than captured by this sketch for it is likely that appreciable diffusion and incorporation of outsiders was taking place in those days, migrations are often back and forth movements, the original ‘urheimat’ might have been depopulated, and any migrating group would be genetically heterogeneous to some extent. See the genebase tutorial for a more detailed writeup by them. The following summary is very vaguely based on a recent paper and the ISOGG page on the R haplogroup, see that page for more up to date information.
A further set of mutations, L146/M420/PF6229 that arose somewhere on the north-south corridor between the Indus Valley and the Eurasian Steppes marks as being R1a (FTDNA calls it R-M417, but that conflicts ISOGG; M343/PF6242 is the marker for R1b); after which M459/PF6235 marks us as belonging to group R1a1; and I belong to R1a1a (also called R-M17) defined by the markers M512/PF6239. I also have the markers M417 R1a1a1. I have not been tested for the SNPs defining the new R1a1a1a (L664/CTS7083/S298) or the new R1a1a1b (S224/Z645) clades Some of my close STR12 matches are from R1a1a1b2a1 L657.1/S347.1 subclade of R1a1a1b2a (F3105/S340/Z94) of R1a1a1b2 (Z93/F992/S202); but one is probably from R1a1a1b1a3b1, and the 1/12 matches are from both R1a1a1b1 and R1a1a1b2 clades. Since the R1a1a1b1/R1a1a1b2 split is probably 12000 years back, finding STR12 matches in both is probably a case of convergence, or we have misdated the split. A complete list of the results of my test is available here: but referring to the outdated 2010 tree.
The R1a1 group probably arose about 20 thousand years ago around the Hindukush mountains (or in Pakistan), and the depth of R1a1a is about 15–16 thousand years in the Pakistan-West India region. From there it may have spread across Nepal and India and to the north-east, reaching the Caucasus by about 12 thousand years ago. From there it seems to have spread to the Southwest through Turkey to Hungary and Greece, the Southeast to Iran, and northwards through Russia. Italy and Kyrgyzstan may have been the last major expansions into fresh territories probably 5.9 and 5.6 thousand years ago. (The eastern European expansion is of the R1a1a1b1 (recently R1a1a1g and, before that, R1a1a7), which arose in Poland about 11 thousand years back, and started spreading by about 8500 years back.) Meanwhile R1a1a1b2 defined by the Z93 mutation probably arose just outside India about 12 thousand years back, and R1a1a1b2a defined by Z94 near the Indian subcontinent probably more than 6000 years back. The history gets murky at this point: it is also possible that Z93 and Z94 arose as far north as the Sintashta culture; L342.2 defining the R1a1a1b2a1 may have spread further among the Andronovo culture, whereas the L342.2- was confined to the Srubna culture in Turkmenistan. In any case, the Z2123 mutation defining R1a1a1b2a1b and Z2124 mutation defining an as yet unnamed clade arose amongst the L342.2+, and probably spread east, west and south from the Indian northwest by about 4000 years back. In this young chronology, L657 defining the R1a1a1b2a1a probably arose on the eastward migration of Z94 people within India, and Y7 defining R1a1a1b2a1a1 arose amongst them around 500 B.C; the Siberian occurences being 16th century admixtures. Alternatively, these groups could have arisen among the Andronovo culture and moved relatively late to India, and were almost lost elsewhere, except isolated instances in Siberia.
These estimates that try to capture isolated tribal populations as well as the more populous groups also use a slow mutation rate estimate, so they are at the older end. The more common estimates use a faster mutation rate and time the major expansions of these groups and typically yield dates about half as old. For example, as illustrated by the maps made by an user called Igor, the major R1a* descendants are found in Europe, who separated from the other branches at about 9000 years back. The R1a1 can then be timed to about 8000–9000 years, and the R1a1a 7200±800 years. R1a1a1b1, the major European and North Eurasian branch, and R1a1a1b2, the South-Eastern branch (including Kyrgyz and Ashkenazi), would both date to 5500±600. R1a1a1b2a1a is the most represented Indian branch that would be dated to 4200±450 years. The European branch splits into Central European R1a1a1b1a1 (3100±350 years, containing the Western Slavic R1a1a1b1a1a at 2775±400 years), Scandinavian R1a1a1b1a3 (5200±600 years) and the rest R1a1a1b1a2 (5000±600 years).
There is some reason to believe that the R1a were the ancestors of the people who spread the Indoeuropean group of languages into Asia, and possibly the Steppes people that domesticated the horse. Of course, if, as seems likely, the R1a haplotype arose thousands of years before the origin of the Indoeuropean group of languages, it probably was present indigenously in various places including India before the advent of Indoeuropean speaking people. Conversely, this late in human history in such a populated region, the amount of linguistic conversion and cultural adoption is expected to be great enough that the people speaking predominantly Indoeuropean languages could often harbour a mixture of haplogroups, and ‘Indoeuropean migrations’ in various regions could even be predominantly represented by different haplotypes. In fact, recent work published shows that the history of the R1a1 people is probably far more complicated than a simple migration of Indoeuropean people either to or from India.