GENERAL ARTICLES CURRENT SCIENCE, VOL. 114, NO. 5, 10 MARCH 2018 971 Partha P. Majumder is in the National Institute of Biomedical Genomics, Kalyani 741 251, India and Indian Statistical Institute, Kolkata 700 108, India. e-mail: ppm1@nibmg.ac.in Understanding the Aryan debate: population genetic concepts and frameworks Partha P. Majumder A long-standing debate on whether ‘Aryans’ (central Asians) had entered India has recently gained momentum. The debate is polarized. In the recent set of articles, some authors have strongly criti- cized inferences drawn using genomic data and population genetic methods. Some criticisms are flawed. These criticisms stem from lack of clear understanding of population genetic concepts and frameworks. Such inappropriate criticisms have led to unnecessary confusion and further polariza- tion among readers. This article attempts to place the ongoing ‘Aryan debate’ in the perspective of population genetic frameworks. Keywords: Ancestry, admixture, central Asia, haplogroup, lineage. SPARKED by a recent scientific publication in BMC Evolutionary Biology by Silva et al. 1 , we have witnessed in the pages of The Hindu (Joseph 2 ; Danino’s critique of Joseph 3 ; Joseph’s rebuttal 4 ) and elsewhere 5 a bitter battle pertaining to the ‘Aryan issue’. The battle will persist. The Aryan issue has been debated with emotion, when it does not really have to be. Population genetic inferences have been criticized as subjective, when it should be con- sidered as more objective than inferences of most other empirical sciences, especially because (a) the framework and methodologies of drawing inferences from data are defined with mathematical and statistical rigour in popu- lation genetics, and (b) better sampling of populations and generation of larger data sets are making the infer- ences more reliable and robust. Irrelevant issues, such as whether there was ‘out-of-India’ migration, are brought into the Aryan debate that adds more haze than transpar- ency. Of course, the ‘out-of-India’ issue is related to the debate on whether Sanskrit is the mother of all Indo- Aryan languages, and moved from India to Eurasia. The fact that many words of Sanskrit are phonetically similar to those in Greek, Latin, Gothic, Celtic, etc. indicates that these languages share a common origin, possibly devel- oped by a set of people now extinct. Sir William Jones, the founder of the Asiatic Society, had pointed this out long ago. The ‘debate’ will, of course, live on, unless there is a strong intent to weigh evidence objectively. In the articles published in The Hindu and elsewhere, many methodological and conceptual issues related to analyses of genetic data of populations have been touched upon. However, these issues were not discussed in suffi- cient detail for the general reader to clearly comprehend the framework and the basis of inferences in population genetics. Even I, a practising population geneticist work- ing on ethnic populations of India for about 40 years, had a hard time understanding the logical veracity of some of the claims made in these articles. Sweeping statements such as ‘most studies of population genetics suffer from shortcomings and flaws…’ 3 have been made. Here, I am neither deeply focusing on the ‘Aryan debate’ nor intend- ing to critique any of these recent articles. My intent is only to try and more clearly inform the reader on certain relevant concepts and frameworks that underlie the popu- lation genetic studies pertinent to these articles. I am dri- ven to do so because the Joseph and Danino essays are being discussed by students and researchers, often with- out clarity of their own thought. I must also point out that population geneticists have not always been able to peg their inferences in the context of social history of a re- gion; sometimes this is not possible; sometimes geneti- cists are not adequately informed of social history. This inability has contributed to a lot of interpretative confu- sion. I will try to provide some clarity to four population genetic concepts that underlie the debate between Joseph and Danino, and also present sketches of relevant inferen- tial methods for a broad understanding. These include, (a) Evolution of populations, (b) Sampling for unbiased and representative genetic information, (c) Inferences from fossil or other palaeontological or archaeological records, and (d) Admixture detection and estimation. Evolution of populations The kind of population genetic studies referred to in the recent articles pertain to evolution of populations. Popu- lations evolve over time. In the remote past, when we