am 19. Mai 2012
This brochure is a concise overview of some of the central themes of Graham Harman's objectoriented
philosophy, in a bilingual English-German edition. The English text occupies just eleven
and a half pages (p4-15). The content is quite engaging as he manages to expound his ideas in the
form of a response to Sir Arthur Eddington's famous two table argument, which can be found in his
book THE NATURE OF THE PHYSICAL WORLD (available as a free download at archive.org),
first published in 1928. This allows him to couch his arguments in terms of a running engagement
with reductionism, in what Harman sees as its humanistic and scientistic forms. So far so good.
Problems arise however when we consider his account of each of Eddington's two tables, and even
more so with his presentation of the eponymous "third table".
1) THE FIRST TABLE: On Analogical Typology
THE THIRD TABLE does not give us a close reading of Eddington's 1928 'Introduction', nor does
it intend to. It's stake lies elsewhere, in a typological reading. From the 'Introduction' Harman
extracts two types, the everyday table and the scientific one, and he immediately puts them in
parallel with C.P. Snow's two cultures, the sciences and the humanities, 'distinguishing', according
to Harman (p5), 'so-called literary intellectuals from natural scientists'. (One may ask why the
literary intellectuals merit the appellation 'so-called', but not the natural scientists. There would
seem to be a residue of scientism that creeps into Harman's exposition from time to time, even
when he is describing his two adversaries).
This is a bold leap of the analogical imagination, but unfortunately in this case it leads him astray.
For the everyday table is not the table of the humanities (whatever that is!). For example, Proust's
table, a locus of intense and meaningful events, is not the everyday table. I am obliged here to
hypothesise that literature belongs to the humanities, as Harman distinguishes the humanities from
the arts, which are capable of rising to, or descending to (depending on your starting point), the
Harmanian table, the 'only real' table. He further complicates the matter by talking about the table
being reducible 'upward to a series of effects on people and other things' (p6). Now this is most
strange, as Eddington insists on the substantiality of the everyday table, its solidity and compaction.
The familiar table is all of a piece, and certainly not a 'series of effects', even less a series of
'table-effects on humans' (p7). So the first table, the familiar one, is a blur between the everyday
table, the table of the humanities, and the 'series of effects' ' with Harman sliding glibly from one
to the other as if it were all the same to him. What seems to guide some of the slippage is the needs
of the analogical argument, now favorising one term in the typological couple, now another. But a
basic incertitude reigns as to the identity of the first table, at least in Harman's text (for Eddington,
as we have seen, there is no uncertainty).
Harman has a real architectonic ambition, and, if only for this, he deserves our admiration and
encouragement. But the typological imagination, with its bold analogical heuristics, can sometimes
play tricks on even the best of us.
2) THE SECOND TABLE: On Reduction
The problem is that Harman seems to have no clear idea of what reduction is. In effect, he presents
us with an epistemological straw man supposed to exemplify the reductionism of modern physics.
While ostensibly talking about Eddington's parable of the two tables, Harman condemns the
procedure of the 'scientist' who, according to him, 'reduces the table downward to tiny particles
invisible to the eye' (THE THIRD TABLE, p6), 'dissolved into rushing electric charges and other
tiny elements'. He contrasts this obviously unsatisfactory procedure of reduction with the OOP's
respect for 'the autonomous reality' of the table 'over and above its causal components' (p7-8). He
informs us that the table is an emergent whole which 'has features that its various component
particles do not have in isolation' (p7).
This is an important point to make, but certainly not to Eddington or to any other physicist worthy
of the name. Perhaps Harman is thinking in fact of Badiou and his set-theoretic reductionism, as he
further declares that 'objects are not just sets of atoms' (p8). However, for any real physicist a table
is an emergent structure of particles and fields of force (not just electromagnetic but also
gravitational and those of the weak and strong forces) and space-time. Even Eddington speaks of
the table as composed of 'space pervaded ' by fields of force', 'electric charges rushing about
with great speed'. Harman is wrong, in my opinion, to treat these 'electric charges' as if they were
just particles, and he pays no attention to the mention of speed. True, Eddington does talk as well of
'electric particles', but there is a progression in the text over the notion of these particles, from
which he first removes all substance (p.xvi), and which he then terms 'nuclei of electric force'
(p.xvii), to finally declare the notion of a particle, such as an electron, too coloured by concretistic
picture thinking and needing to be replaced by mathematical symbolism:
'I can well understand that the younger minds are finding these pictures too concrete and are
striving to construct the world out of Hamiltonian functions and symbols so far removed from
human preconception that they do not even obey the laws of orthodox arithmetic' (p.xviii).
Thus, contrary to what Harman asserts, there is no 'reduction to tiny particles', but a redescription
in terms of a complex, emergent, structure of forces and fields and regions of space-time.
I think Harman confuses reduction between different worlds with reduction inside a particular
world. If scientists declared that the physicist's table was the only real table, as Harman does with
his philosophical table (he calls his third table, which can neither be known nor touched, 'the only
real one') then that would be a form of reductionism. But we have seen that there is no reduction of
the table to a set of tiny particles (how big is a field of force? how far does it extend? Harman is so
obsessed with refuting a non-existent particle-reductionism that he does not consider these
questions, and goes on to protest against an imaginary 'prejudice' that maintains that 'only the
smallest things are real' (p8). This is precisely the picture-thinking that Eddington is eager to dispel
in the physicist's world). There is no 'disintegrating' of the table into nothing but tiny electric
charges or material flickerings' (p10). There is no 'scientific dissolution' (p8) of the table into its
component atoms, as this would be merely be bad science. To this extent, Harman's new objectoriented
ontology is just bad epistemology.
3) THE THIRD TABLE: Tabula Trifecta
A) LOST IN SPACE: can't tell up from down
We have seen that Harman's confusion over the nature of reduction leads him to misrepresent
Eddington's explication of the physicist's table. Adam Robbert proposes to import into the text the
notions of 'undermining' and 'overmining'. But I don't think that this can help. Harman has
centered THE THIRD TABLE on a critique of the two tables described by Eddington, as products
of 'reduction downward' (the physicist's table) and of 'reduction upward' (the everyday table)
respectively. These terms seem to be the equivalents of undermining and overmining in the text, but
I can't be sure. Adam Robbert seems to think that reducing the table to particles (which Eddington
does not) would correspond to undermining, and reducing it to electromagnetic fields (neither does
he do this) would be overmining. But surely the size of the components of an object cannot count as
the demarcation criterion of undermining vs overmining. If an object is reduced to its component
parts, this should be called undermining, even if the components are bigger than the whole they
compose. A physicist would be guilty of overmining, on this interpretative hypothesis, if he reduced
the table to a space-time slice of an all-embracing set of fields (which, once again, Eddington does
B) Ethics and objects: love the table
Harman situates his third table in an epstemological space, under the everyday table (or is it the
humanist's table? or the bundle-of-effects-on-humans table? Given the three variants of the first
table perhaps we should talk in terms of table1-1-, table1-2, and table1-3)
and above the scientific table:
'By locating the third table (and to repeat, this is the only real table) in a space between the 'table'
as particles and the 'table' in its effects on humans [i.e. table-13], we have apparently found a table
that can be verified in no way at all ' Yes, and that is precisely the point' (p11-12) The table is not
just untouchable, it is also unverifiable, which in Harman's epistemology seems to mean that it is
unknowable, as he adds
'The real is something that cannot be known, only loved' (p12).
Loving the table is to be understood in terms of the 'erotic model' (15), whose principal feature
seems to be indirectness and obliquity. Many have talked about reinstating an erotic knowledge of
objects, of unifying eros and logos. I am all in favour of this idea, and I think the work of the
Scottish poet Kenneth White is gives us many useful indications of what such an erotic model
entails. But these thinkers typically advocate a sensual encounter with objects in their singularity,
and Harman emphasises rather the non-sensual withdrawal of his intelligible objects. For they are
intelligible and not sensible, not touchable, not verifiable, not knowable. We are deep in negative
theology country here, far from the familiar objects of everyday life.