Casta is not a term indigenous to the Indian subcontinent.  It is in fact a Portuguese term, which was used mostly in the context of Iberian colonization of South America, to describe what were thought to be different racial groups: blacks, Indians, Europeans, creoles and so on.   One interesting feature of this usage was the way in which different admixtures between recognized races were identified as part of an elaborate racial taxonomy, as represented in so-called casta paintings (Carrera, 2003).   As Carrera notes, this approach owed much to the development of Enlightenment thinking in Europe, in which systems of classification, developed as a feature of natural science, were also applied to the interpretation of social groups.  There is some evidence that a similar approach was deployed by Portuguese travelers in India to identify ‘classes of heathen’ such as ‘Bramenes’ (for example, by the early sixteenth century traveller Duarte Barbosa, as quoted by Dirks, 2001: 19; see also Dumont, 1980: 21), although there is some debate over the extent to which such early travelers actually focused on this idea as a fundamental feature of Indian culture.  For our purposes, however, the reference to the Portuguese root of the English term ‘caste’ is significant primarily because it places emphasis on the emergence of this term through the general encounter between European and non-European peoples during the age of expansion.  As with casta taxonomies in colonial South America, caste in India developed into a kind of scientistic key through which European powers acquired knowledge about their subjects.  Caste came to be seen as a kind of ‘natural’ feature of Indian society, a fundamental, even timeless form of social organisation (Inden, 1986: 427-9).

In a key intervention, the historian Nicholas Dirks has argued that it was this principle which underpinned the development of the so-called ‘caste system’ during the colonial era. ‘Caste (as we know it today),’ he says, ‘is not in fact some unchanged survival of ancient India, not some single system that reflects a core civilisational value, not a basic expression of Indian tradition.  Rather (it) is a modern phenomenon, …the product of an historical encounter between Indiaand Western colonial rule’ (2001: 5).   This looks rather as if it might be claiming that caste is a nineteenth century invention, as some scholars have argued of ‘Hinduism’ itself (for a survey and critique, see Pennington, 2005: 169-72).  Dirks, however, is careful to argue not that caste was invented, but rather that it was ‘systematized’ during this period.  This process of systematization occurred partly through the progressive naturalization of caste in Orientalist and other writings about India, and through the development of what Dirks calls the ‘ethnographic state’. He makes the case that, particularly in the second half of the nineteenth century after the famous rebellion of 1857, the colonial state was geared towards gathering information about its subjects with a new vigour, frequently analyzing social and political relationships in a particular area according to the perceived characteristics of different castes.  This tendency was reinforced by the census (Ibbetson, 1883; Raheja, 1996; Hann, 2005).  Just as the census had a profound impact on the perception of religious community in India, so,  Dirks argues, a similar process of reification occurred in relation to caste.  This entrenched the idea that caste constituted a complex yet coherent system of hierarchical organisation which governed social life.  It is on this basis, he continues, that caste has emerged as a major feature of post-colonial political life.  Such a process also, he argues, forms the basis for the development of academic approaches to caste.  In particular, he critiques a very famous 1960s text about caste by the French anthropologist Louis Dumont, called Homo Hierarchicus.  Dumont’s key thesis, which has been critiqued from a number of angles (see Appadurai, 1986, 1988; Quigley, 1999, chs. 2 & 3), was that caste formed a structuralist rationale underlying Indian society, in which religiously (that is, Hindu-) inspired notions of purity and impurity configured the position of all individuals, relative to an overall system.  The resulting hierarchy, he argued, was a ‘natural’ tendency of social organisation (unlike the ‘artificial’ tendency towards equality and individualism he noted in the modern West); the particularity of the Indian context was that power was subordinated to status in this hierarchy, as ritual purity was more significant than political status. Although Dumont argued that this did not preclude dynamism, in the movement of caste groups relative to one another, that dynamism was seen as part of the internal logic of the system, governed by the purity/impurity rationale.

The systemic character of caste-as-casta is, then, one which clearly has specific historical roots. Although we must take account of it, because it has had a major impact on the operation of caste in the modern world, we also need to analyse it in terms of broader contexts, interrelated ideologies and social practices which operate differently in different situations.  Undeniably, this makes understanding caste more complex.  Two further key concepts will help us to get to grips with this complexity: jati, and first of all, varna.