In Kenya, violence abounds, as do analyses of its causes and consequences. An efficient way of dividing opinion on the matter is to ask four questions: Was the violence planned? Was it 'ethnic'? Was there ethnic cleansing? Was it 'political'?

We aren't short of people who will answer no to all save the last for boring political reasons; we needn't worry about them. But others, for presumably non-political reasons, will do likewise. 'Unless names are invidiously named', as Timothy Williamson once said, 'sermons like this... tend to cause less offence than they should, because everyone imagines that they are aimed at other people'. Maina Kiai, Tavia Nyong'o, here's looking at you. 

My targets offer distinct arguments. Tavia argues that ethnicity is the form, rather than the content, of Kenyan politics: 

How does ethnicity factor into Kenyan politics? Beyond the English and Kiswahili-speaking elites, it certainly is the form of politics. How and when politicians address audiences in their native tongues is a constant preoccupation for Kenyans, who have gone so far as to require that all leaders speak Kiswahili, the only language Kenyans consider to have transcended the nation's ethnic distinctions. The hardened political identities that the scholar Paul Gilroy has described as "encamped" can emerge in periods of stress, and as the worst crisis since the 1982 coup has unfolded, Kenya has indeed seen those hardened identities unleashed. But most Kenyans recognize and reject tribalism wherever they see it. Ironically, the big push to unseat President Kibaki was largely motivated by this very rejection of tribalism; during his five years of power, he has favored a tight clique of drinking buddies from the same few villages near Mount Kenya. And amidst the horror there are signs of promise, like the NGO Common Hope for Health, which is delivering aid across tribal lines and defusing tensions between rival youth gangs in the west. The vast majority of Kenyans still long for peace and a shared homeland.

Ethnicity is the form of Kenyan politics, but the content is the same as everywhere else: money, power, land, influence, and, increasingly these days, democracy. Opinion was divided over the wisdom of the opposition protest rally called for last week in Nairobi's Uhuru Park. But a fear that Kenyans could not conduct open, democratic politics in a multicultural context did not stop all the major presidential candidates from holding rallies and political events in the lead-up to the largely peaceful and orderly election last Thursday. Kenya is not Rwanda, and reports of a "majority" and "minority" tribe engaged in a struggle to the death are widely off the mark. In reality, all of Kenya's leaders know that in order to govern the country they must put together a coalition of interest groups from many regions. It is even written into the constitution that any successful candidate for president must poll at least 25% of the vote from a majority of Kenya's provinces.

Maina's testimony before the US House of Representatives' subcommittee on African affairs included the claims that:

 
The violence is neither genocide nor ethnic cleansing: The root of the problem is not that different ethnic groups decided they could no longer live together. The root of the problem is the inability of peaceful means to address grievances. For this to be genocide there would have to be either state complicity or state collapse and the first obligation would be for the state to provide adequate security for those at risk. Instead we have uneven and selective policing with emphasis on preventing Raila Odinga from holding protests in Nairobi rather than protecting IDPs and others at risk across the country. We therefore believe that the quickest and most effective way to reduce the violence is progress in the current talks.

 
I hope it is evident that the reasoning in each case is unsound. The notions of 'form' and 'substance', which are central to Tavia's piece, are nowhere clearly characterized. If the substance of politics is the goal at which political action is aimed, then the central claim of the article is false. Kalenjin ethnonationalism, at least in its extreme reaches, is territorial: one of its central aims is the exclusive enjoyment of large parts of the Rift Valley. And it imposes normative requirements: Seroney was lauded for the Nandi Hills Declaration in 1969; then, as now, Moi was blamed for not restoring ancestral land to Kalenjin ownership. On a natural construal of 'substance', ethnicity is part of the substance of Kenyan politics. 

Maina's testimony may be disposed of briefly.

Neither state complicity nor participation is necessary for a legal finding of genocide. The International Criminal Tribunal found that the Srebrenica massacre was an act of genocide, and that while Serbia was not directly responsible for the massacre it also failed in its duty to prevent genocide. Complicity requires positive action in aid of the commission of a crime, so the finding of negligence in preventing genocide does not guarantee that Serbia was complicit in the massacre. These findings appear to constitute a counterexample to the claims given in Kiai's testimony.

It is not, anyway, a condition on either genocide or ethnic cleansing that whole communities (or even majorities thereof) decide that they cannot live with each other: by that standard, there was no genocide in Rwanda in 1994, for many Hutus chose to defend Tutsis. Equally, from the fact that Kalenjin who protected Gikuyu or refused to participate in the violence were often assaulted, it doesn't follow that ethnic cleansing didn't happen in RVP. Likewise, of course, for events in Nyanza, Central Province, and elsewhere. The criteria given are transparently irrelevant. 

So, we have obviously clever people making obviously false assertions and giving obviously bad arguments for them. What gives? We'll get there. First, a quick reminder of the lie of the argumentative land.

In the past, the Luo Union traded on exclusively ethnic lines; in the present, Transcentury has precisely zero non-GEMA shareholders.[1] Months ago, I had the utterly surreal experience of reading, in quick succession in a bookshop in Nairobi, Colin Leys lament the inability of the Kenyan middle classes to overcome ethnic cleavage in their own obvious common class interests, and Prof. Peter Anyang' Nyong'o on the necessity of breaking off the Gikuyu bourgeoisie (in an obituary for Ramogi Achieng' Oneko, no less).[2][3] Ethnicity already motivates this kind of irrational behaviour, we should expect it to motivate violence too.

f you want to explain the violence, you need to explain how attackers are mobilized, how victims are chosen, and how attackers conceived of their victims (more precisely: why the attacker thought the victim merited the attack). Class is a poor explanation for all of those: for much of the post-election violence, it doesn't determine as neat an explanation as does ethnicity. But wait, why are we talking about class and ethnicity at all? Probably because class and ethnicity are (pre)supposed to be exclusive and jointly exhaustive explanations for the violence. The fight is then over which of these is the real cause of the violence.

The offending assumptions come in different flavours: the unsophisticated idea that class and ethnicity are competing explanations and that class is the sole proprietor of the causal role, and the slightly more sophisticated thought that even if class and ethnicity jointly explain the violence, class is the ultimate explanation because ethnicity is reducible to it in some way. Pending the promised submersion of ethnicity into class, we have no compelling reason to accept the second. The first ought not to be taken seriously: manifestly, there have been distinct kinds of violence; those who wish to show that there is one cause for all of them bear the burden of proof; it is a heavy one.

To show that the violence is politically motivated is to make it respectable (or at least comprehensible). Class is a vaguely 'political' concept, so the violence is explained in class terms. The gain in intelligibility justifies the choice of class to explain the violence. That overlooks the point that killing people for the identities they bear is also a political act. Any intelligibility gained by appealing to political categories is as available to those who choose identity as it is to those who choose class. Respectability, it appears, is the only reason for denying ethnicity its explanatory power.

Returning to Kiai and Nyong'o, you'll notice that they're at one in denying that ethnicity has independent motivating power: for them, its potency is derivative. In this boggy and treacherous terrain, that thought is the landmark by which the parties orient themselves. The premiss is false, which is why those who take their bearings by it have arrived at such strange conclusions. Ethnically-motivated violence is what it is and not another thing. The beast stares us in the face. Time to stare it down.

[1] See B.A. Ogot (2002), "Luo Identity and History" in The Challenges of Leadership and History in Africa: The essays of Bethwell Alan Ogot for the exclusivity of the Luo Union

[2] Leys, p. 161 of The Rise and Fall of Development Theory.

[3] Nyong'o, (2007) "Profiles in Courage: Ramogi Achieng Oneko", in A Leap into the Future : A Vision for Kenya's Socio-Political and Economic Transformation. African Research and Resource Forum: Nairobi