Machine learning — getting results that are completely wrong - from Lars Syll Machine-learning techniques used by thousands of scientists to analyse data are producing results that are misleading and often completely w...
Wednesday, 11 July 2012
It starts with an analysis of different modes of exchange, starting from the bottom up, and then briefly considers the nature of value (but this is a topic left mainly to a previous Graeber book). It then starts on its main theses, questioning some long held assumptions about exchange, notably that before monetary systems there was barter. It argues (using anthropological cases) that informal systems of mutual obligation preceded money and barter, if it occurs, does so in significant amounts, only in cases where there was a monetary system that has collapsed. Rather it contrasts these social systems, which are primarily to organise human relations (marriage, honour, reparation, status etc.), with that of accounting and debt which it sees as coming out of military power and its needs. The former case is made with much greater persuasiveness than the later contrast, which seems ideologically driven and not clearly illustrated by the examples given.
Although it finally shies away from definite conclusions, it implies that formally accountable debt always ends in collapse and (effectively or actually) slavery, and hence has to be wiped clean every now and then. It is somewhat utopian in that paints the informal systems of social obligation in a rosey light and taints formal systems by association with slavery and militarism without proving a causal connection. The causation is somewhat naive, in the sense that grand macroscale narratives are drawn, and systemic advantages of accounting systems ignored. Indeed it tends to ignore the more mundane processes of the everyday market for these grander concerns, implying that insisting on debts being repaid is socially arbitrary, without the effects upon a system of exchange of this really being considered. This reveals the weakness of the purely anthropological approach, in contrast to its effective undermining of "received" assumptions of how things have to work.
The books strengths are undoubtedly the fresh and different viewpoint, informed by numerous anthropological examples. However the weaknesses of the anarchist narrative that emerges are slowly revealed as the book slowly unravels as it progresses. Grand narratives of coordinated epochs across the whole world, including civilisations that barely communicated stretch credulity to the limit. However the shaking up of thought and the freeing from a tired monetary vision of exchange is very refreshing and hence welcome.
A more systematic synopis of the book can be found at: http://en.wikipedia.org/wiki/Debt:_The_First_5000_Years, and his academic home page is at: http://www.gold.ac.uk/anthropology/staff/d-graeber/