Breaking down the question
The question asks you to analyse the differences between two ways of studying caste — the attributional and the interactional approaches. This is a comparative analytical task: you must define each approach, then draw out the axes on which they diverge, ideally with named exponents and illustrations.
Analyse means more than list. You should organise the differences along clear dimensions — the unit of analysis, how hierarchy is determined, the method, and the underlying image of caste — so that the contrast is systematic rather than piecemeal.
How to approach it
Begin by defining both approaches in a sentence each. Then structure the comparison around a few axes: what determines a caste's rank, whether caste is studied in isolation or in relation, static versus dynamic, and the method used. Attach thinkers — Bougle, Hutton and Dumont to the attributional side; Marriott and the interactional field-workers to the other. Close by noting they are complementary rather than mutually exclusive.
Model answer
Sociologists have studied the caste system through two broad lenses. The attributional approach explains a caste's position by its intrinsic attributes — the ritual qualities, occupation and customs it possesses. The interactional approach explains position through the relations and transactions between castes — who accepts food or water from whom, and who defers to whom in practice.
The differences may be analysed along several axes.
Basis of ranking. The attributional approach ranks a caste by the attributes it holds — degree of ritual purity, vegetarianism, occupation, endogamy and lifestyle. Emile Senart, Bougle, J.H. Hutton and, most influentially, Louis Dumont built on this, Dumont grounding the whole system in the single principle of purity and pollution and the opposition of the pure and the impure. The interactional approach, by contrast, holds that rank cannot be read off attributes alone; it is revealed only in interaction — in the patterns of giving and receiving that place castes above or below one another in actual village life.
Unit and image of caste. The attributional view tends to treat each caste as a bounded entity with fixed characteristics, yielding a relatively static picture of hierarchy. The interactional view, developed through fieldwork by McKim Marriott and others, sees rank as relational and dynamic — it can vary locally, is constantly negotiated, and is best mapped through transactional networks. Marriott's work on food transactions and matrix rankings exemplifies this.
Method. The attributional approach draws heavily on texts, varna theory and general classification. The interactional approach is rooted in empirical fieldwork and observation of concrete behaviour, a legacy of the village studies of the 1950s.
Explanatory scope. Attributional accounts capture the ideological, pan-Indian logic of caste. Interactional accounts capture its local, lived and often inconsistent operation — explaining, for instance, how a caste's ritual attributes may not match its actual dominance, as with Srinivas's notion of the dominant caste, where secular power shapes standing.
The two are best seen as complementary. Attributes supply the cultural code; interaction shows how that code is enacted and sometimes subverted on the ground. For deeper treatment, see the note on perspectives on caste.
Examiner's perspective
Examiners reward a genuinely analytical comparison — differences organised along explicit axes such as basis of ranking, unit of study and method — rather than two separate descriptions. Correctly attributing the attributional line to Dumont and Hutton and the interactional line to Marriott shows command of the literature. The strongest scripts resist treating the approaches as rivals and conclude that they illuminate different faces of caste, the ideological and the interactional.