Tax or tribute?
The problem of what sources say and what they mean
I promised you one last post in this series about a poem from the early first-millennium CE poetry collection, Puranānūru, which I first shared four weeks ago. I’ll copy it again here because I want to home in on one line:
As the hostile onslaught bent on destruction beats at the front ranks
of the army of his king, he alone, sword raised in his right hand,
blocks the enemy from surging on and so he is like a shore
to a great ocean, that man whose lineage, like him, has shown
preeminent generosity, the ruler of a village which other than always
feeding those who come to it singing
does not even create enough income to merit being taxed by a king.
The song of Maturaik Kanakkāyanār.1
The line I want to focus on is the last one, ‘[whose village] does not even create enough income to merit being taxed by a king’.
When I read that line, my reaction was, ‘Wait a minute, I don’t think somebody could have been paying tax on income in first-millennium CE southern peninsular South Asia’.
I think this is an example of where what we know historically about a society changes how we understand sources from it. Of course, what we know historically about a society also comes entirely from those sources. Building up a picture from as many different sources, and as many different kinds of sources, as possible is vital because it sets off exactly this kind of dialogue: the more we know, the more accurately we can see the materials that we use to know, and so the more we can know.
Surely, though, if it says ‘income’ and ‘tax’, then that is what the source says and we just have to work with that?
Well, yes… and no.
If a source says something (or depicts something, or archaeologists dig something up), that is what it is. Assuming we are sure it has really been linked to the right date and place, then we have to fit it into our picture. Our picture can go beyond the sources, by using comparison, analogy, logic, etc., but the picture we create should never be rule out sources that we do have.
But that does not mean that interpreting the sources is simple. As I said last week, the language used in ancient Tamil poetic sources is extremely complicated. However, even when we know exactly how all the grammar fits, what exactly did people mean?
Words change their meaning all the time. People actually living in a society, instead of trying to explain it to outsiders, or historians centuries later, use slang, jokes, abbreviations, set phrases and metaphors which make total sense to them.
And then there is the issue of translation between languages. If I say that the source says ‘income’ and ‘tax’, I’m using two English words with their own massive, complicated and debated meanings, to stand in as translations for two Tamil words which also have their own internal histories. In later texts they can indeed mean things related to tax and income, but these are the earliest records we have of them, so it is hard to know if they started out that way.
How else could we translate those words?
One option is ‘produce’ and ‘tribute’. Another is ‘stuff’ (in the sense of food and goods) and ‘loot’.
What happens to that line if we use each of these options?
‘[whose village] does not even create enough produce to merit being asked for tribute by a king’
‘[whose village] does not even create enough stuff to merit being looted by a king’
Each of these versions paints a very different picture of what life looked like for both the villagers and the king.
Tax and income
In the version talking about income and tax, we have to imagine a bunch of other things to make sense of those words.
Tax is a regular, specific demand made by a government or ruler, usually calculated to be a particular portion of production or to be distributed evenly over particular areas of land or population. In other words, it is designed to appear transparent, predictable and fair (even if it was often none of those things and even if people frequently complained about this). And that means that tax needs other things:
a system of evaluation: at minimum, to levy taxes a government or ruler needs to know how many people they have to tax and/or how much land they have to tax. It helps to know lots of other things like how good the land is, how many of the people are of working age, how many plough animals they have, and so on.
a system of record: it is no use having a system for evaluating how much tax a place for a person or a community pays without a way of recording that and keeping those records at least somewhat up-to-date. It is also pretty critical to be able to record who/where has or hasn’t paid their taxes. And, if a government is going to use tax breaks to reward people (which they often do and did) that also needs to be recorded.
a system of collection: there’s a reason they say only two things are inevitable, death and taxes, and it isn’t because it’s true. Lots of people across history will have lived without ever paying taxes, but they will all have died. It’s a popular saying because it expresses something else about taxes: that lots of people don’t much like them (and, in fairness, I think the saying originated in the US, where tax is lot more normal than it has been in lots of other places). Still, the point is, people mostly don’t like paying taxes and therefore generally won’t without at least a system to make it easy (like somebody coming to your village, specifying how much everybody owes and insisting on leaving with it) and, usually, a system for making it hard not to (like that tax collector also being able to call in men who will beat up non-payers and, in extremis, being able to call in the government/ruler’s own army).
If we take seriously the term ‘income’ then we can add an even greater complexity: there would have needed to be a system for evaluating not just what was produced by a community but also how much they sold it for at market, in order to have an ‘income’, which in turn requires markets and probably some unit of value (which we might call ‘money’) so that somebody could calculate total income out of what were probably thousands of different kinds of produce (food, artisanal products, wood and building materials…) and services.
What about,
‘[whose village] does not even create enough produce to merit being asked for tribute by a king’?
Tribute is a term that historians, especially of the Middle Ages and ancient world, often use to describe something that isn’t quite tax. It can be semi-regular, but is usually not something that is expected to be paid by everyone, everywhere, forever. Instead, it is often specific to particular communities, via special relationships with the ruler/government. It also often involves people who are not part of the ruler/government’s own state.
Foreigners could be expected to pay tribute, and indeed, in many medieval and ancient states, the system was exactly that citizens paid tax and outsiders paid tribute, for example, for the privilege of not being invaded. It could also be taken from people within a state, though, and in this case can be considered different from tax in that it was not necessarily intended to be fair - one community or people could pay more tribute than another, or any tribute when others paid none, for all sorts of reasons.
Directly related to this, it could be openly punitive, whereas tax is usually framed as being a reasonable or mutually beneficial arrangement. And tribute was often more directly related to the threat of violence: pay tribute or be invaded, pay tribute or have your village robbed and burned. The line between tribute and tax can be very blurry in some contexts, but broadly, its down-the-line bureaucratic requirements are lower. For a tribute system, there needs to be:
somebody with enough armed men to threaten people into paying or physically take the tribute. The plan would be to leave people with enough to carry on producing so that more tribute could be demanded from them in future (see ‘looting’ below, for when this wasn’t true), but beyond that, tribute could be as much as a community could give or as little as the tribute gathers could be bothered to collect.
maybe some sense of where has been hit up for tribute lately, or better still, a regular calendar of where pays tribute when, so as not to exhaust communities or drive them to rebellion.
Finally, let’s think about
‘[whose village] does not even create enough stuff to merit being looted by a king’
What needs to be in place here?
Pretty much nothing.
A man could call himself king, and as long as he had a gang of heavies around him numerous or mean enough to cause some farmers to back down, he could wander with them, or set himself up somewhere and send them from village to village, taking as much as they could find and carry. Looting is not fair, not transparent, not routine, regularised or particularly sustainable. It isn’t perhaps a very longterm solution, but we shouldn’t be too naive about that. Even if people came through and looted a village every few years, people might still rebuild, re-plant their crops and try again because what else are they going to do?
People, now and in the past, have had to live with temporary situations (let’s say, ten years or less) and much longer term arrangements, in which people might turn up any time with armed men and take whatever they could. Mostly, they would not have known how long the situation was going to last, and frequently, the only options available, other than starvation, were to dust off and start over, maybe try to hide some produce as soon as it was produced, possibly put together some sort of local militia (which could backfire and lead to much worse violence) or find somewhere to run to in a nearby forest or up a nearby mountain and sit it out (still losing a lot of stuff in the process).
Each of these translations, then, suggests a different picture of the Tamil-speaking regions of South Asia in the early first millennium CE. It is not possible to make the decision on linguistic grounds. The poems say what they say but do not explain their terminology. Instead, we have to choose how we understand those words based on what we know about the society from all of the historical sources available to us.
We can certainly start with the poems themselves.
As we have seen, they are full of details of everyday life, of the affairs of kings and of matters of war. Do all of those details add up to something that looks more like taxin income, taking tribute on produce or looting stuff?
It is, inevitably, a mixed picture, because these are ‘ideal types’ or categories invented to define how one thing is different from another. Reality doesn’t do categories. They are just how we describe it. They are useful but messy around the edges. Still, if we think of reality as having clusters or hot spots, which are what our categories are trying to describe, then we might get somewhere.
There are lots of references in the poems to farmers farming and other people, for example, in the hill and mountain areas or on the coast gathering valuable products like honey or hunting deer and fishing. The poems present an idealised reality, but it is not one that contains any reference to hideaways in the forest or walls and fences around villages. The man in the poem is on a field of battle with his king, not fighting to repel raiders from his village. This does not look like an environment where people expect to be looted by any passing band of men.
There are also lots of references to kings, living in big halls, with large walls, and organising feasts and parades and keeping elephants. These are all difficult, expensive, time- and resource-intensive things to do, which is tough (though not impossible) on a diet of random looting. It is unpredictable, irregular and, if an area becomes exhausted, subject to diminishing returns. Keeping court in this way is much easier with a regular, at least semi-predictable source of income.
There are not, however, references to bureaucrats and administrators, to records or assessments, to specific, measured units of land, to population counts or to money. There aren’t really references to markets, either. Instead, people are depicted swapping different types of goods, or bringing them all to the court of the king, where they would be shared through feasting.
What about in other sources? We looked a couple of weeks ago at hero stones, a very small number of which date from the same period as these poems. They point towards a landscape in which violence, and especially cattle raiding, was quite common. Defending a village from raiders and defending it from tigers seem to have been good reasons to give someone a stone.
None of the sources we have for the first two centuries CE point to anything like a tax taking system in the southern South Asian peninsula. Instead, they suggest a world in which mostly kings took regular, or semi-regular tribute and supplemented this with raids on rivals’ kingdoms, or, looking at it from the point of view of the local populations, looting.
In a few centuries, kingdoms would grow larger and more stable and those regular tribute payments would start to be presented, often in stone inscriptions which also served as a permanent record, as transparent, fair and predictable. Later poems would mention administrators and bureaucrats. But, as far as the totality of evidence can suggest at this time, for the early centuries CE, it was not possible for a village to pay tax on an income.
This does not change much when it comes to the heroism of the poem’s subject or the beauty of its imagery, but it alters how we think about his relationship to his king, what we think the purpose and scale of the battle might have been and exactly how we might translate that final line…
Hart, George L., and Hank Heifetz, eds., The Four Hundred Songs of War and Wisdom: An Anthology of Poems from Classical Tamil, the Purananuru (Columbia University Press, 1999), song no. 330 (p. 188).