Johannesburg — THERE are two very different ways of contemplating a split in the African National Congress (ANC).

For some analysts, the creation of a breakaway party would represent the beginning of normal democratic political life in SA. For others it would signal the end of such normality.

On the first view, the collapse of any dominant party is a cause for celebration. ANC fragmentation is a necessary precursor to the emergence of a competitive party system. A so-called centrist grouping associated with former president Thabo Mbeki could create a party -- let us call it the New ANC -- that appeals to the new middle class and those disinclined to accept the post- Polokwane leadership.

Mbeki secured four votes for every six Jacob Zuma was able to muster, so this breakaway party might begin from a position of considerable strength.

The New ANC would compete for votes against a leftist, trade unionist and "populist" Old ANC dominated by Zuma or his allies.

Eventually, on this rosy account, we might see the two parties alternating in power.

In the fantasy world inhabited by many Democratic Alliance (DA) strategists, moreover, the DA would be able to fulfil its greatest historical yearning. Holding the balance of power between Old and New ANCs, it would supposedly throw its weight behind the centrist grouping and so rescue the country from native misrule.

On the second interpretation of an ANC split, a new party will destroy the relative political stability that a unified and coherent ANC centre has hitherto guaranteed. Although analysts have mostly emphasised the ideological and personal motivations behind Zuma's victory in Polokwane, it has been impossible to overlook what United Democratic Movement leader Bantu Holomisa recently described as the tribalist dimension of Zuma's rise.

Castigating "100% Zulu boy"

T-shirts and ethnic campaign slogans, Holomisa complained that the ANC had been following the "path of tribalism" because Zuma's regional power base in KwaZulu-

Natal was pursuing an "ethnic- driven project".

Holomisa overstates his case, of course. The Polokwane rebellion was fuelled in part by resentment about many decades of disproportionate Xhosa influence in the ANC -- and latterly in the higher levels of the public service. However, the evident ethnic and racial diversity of an executive leadership that has included Essop Pahad, Joel Netshitenzhe, Frank Chikane and Trevor Manuel illustrates the continuing strength of the ANC's antitribalist and nonracist conventions. It also underlines the limitations of any "Xhosa nostra" hypothesis about an ethnic monopoly of power.

While many of Zuma's key allies may be Zulus -- most notably Siphiwe Nyanda, Zweli Mkhize and Blade Nzimande -- many are not.

The Polokwane successes of Kgalema Motlanthe, Gwede Mantashe, Mathews Phosa and Jeremy Cronin demonstrate the continuity of ethnic and regional balance in ANC politics.

Moreover, while political networks centred on Eastern Cape have been undermined, the rise of activists born and raised in Alex and Soweto in the 1940s and 1950s has been as significant a factor in this reconfiguration of power as the resurgence of KwaZulu-Natal.

The threat that ANC fragmentation poses is that it could institutionalise and entrench what might otherwise be passing ethnic tensions.

The New ANC might start out with a high degree of regional and ethnic diversity among its leadership. On the basis of Polokwane nomination and voting patterns, it might secure a substantial vote in Limpopo, Free State and Western Cape. However, it would almost certainly perform disproportionately well in Eastern Cape.

The danger this presents is that the New and the Old ANC would compete for votes in Gauteng, Limpopo, and elsewhere, but that each would possess a regional home that gave it an unmistakable ethnic character.

Antitribalism has been one of the most important attributes of the liberation movement since its formation, and it cannot be safely abandoned today.

Hopefully, the continuing rise of the younger generation of Soweto- and Alexandra-born politicians -- who do not carry with them the dangerous baggage of regional power bases -- will ensure that the disruptive potential of ethnic division will not be realised.

Butler teaches public policy at the University of Cape Town.