TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.1 (283)
chapter
“All things think”
Panpsychism and the metaphysics of nature
Iain Hamilton Grant
Panpsychism has both a negative and a positive value to contemporary metaphysics.
Negatively, it provides a critique of the problem of epistemological and/or phenomenological access as a precondition of metaphysical validity. This critique is pertinent because the precondition is as ubiquitous as it is unacknowledged in both postKantian and post-Humean metaphysics, an assumption that we will call the principle
of finitude. Positively, by setting panpsychism against emergence, it opens a problem at
the heart of contemporary metaphysics of nature, namely, the composition of nature
from powers. Briefly stated, this essay will argue that if nature is so composed, then
reason must be amongst its powers. In so doing, we will note how this composition recasts the panpsychism-emergence problem and removes from the principle of finitude
its authority over reason.
By addressing panpsychism from the perspective of both post-Kantian and postHumean metaphysics, I wish to indicate that these two ‘schools’ share more than they
dispute.1 This becomes especially apparent in the context of the contemporary metaphysics of nature. On the post-Humean side, metaphysicians of nature argue about
the ‘groundedness’ of powers – are they ontologically basic, or properties of a more
basic substance? Some argue for the ‘ungrounded thesis’ (Mumford 2006), and some
hesitate between powers and substances (Molnar 2003). On the post-Kantian side,
metaphysicians argue about the groundedness of reason – what candidates may satisfy the principle of sufficient reason? Some argue that this is best explored through
the metaphysics of nature (Grant 2006), and some through ‘pure reason’ (Meillassoux
2008a). What vitiates all such projects, however, is an attachment to the metaphysics
entailed by what I will call the principle of finitude, which we shall address below.
Before discussing the principle of finitude, I wish to demonstrate why it is that
the problem-field of panpsychism requires address not from the philosophy of mind,
which presupposes the access problem, but rather from the perspectives of ontology,
. That is, by what used to be called the ‘Anglo-American’ and ‘continental’ schools of philosophy.
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.2 (284)
Iain Hamilton Grant
on the one hand, and the philosophy of nature on the other. To do this, I shall derive some salient features of this expanded address to the problem of panpsychism by
contrasting them with ontological claims from which they may be considered to derive.
.
In what does the identity of being and thinking consist?
First amongst these resources is Parmenides’ identity thesis concerning being and
thinking, which establishes the ontological scope of the problem. Fragment 3 reads,
“for thinking and being are the same (to gar auto noein estin te kai einai).” Taken at
face value, the panpsychist implication is immediately evident: whatever is, we might
say, thinks. Yet even reformulating it to this slight extent presents us with problems.
The inference from ‘being is identical to thinking’ to ‘whatever is, thinks’ has added
a qualification to being, now considered as composed of singular beings that are and
that think, rather than being as such. To further clarify the point, consider Cornford’s
account of Parmenides’ identity thesis: “It is the same thing that can be thought and
can be,” which he sets against what he takes to be the patently absurd thesis that “to
think is the same thing as to be” (1964: 34). Behind both these versions lies the assumption of a thinking subject, one that either possesses the capability of thinking what is
(where ‘what is’ is an object the ‘being’ of which satisfies the necessary conditions of
thinkability), or that is by virtue of thinking. This becomes especially clear in Cornford’s
translation of fragment 8: “Thinking and the thought that it is are one and the same”
(1964: 43), which restricts concern to the thought of ‘what is’. But does thinking entail
a thinking subject? No such assumption is evident in either fragment: while fragment
8 asserts the identity of thought (noein) and its object (noema), fragment 3 asserts the
sameness of thinking and being in their infinitive forms. Although therefore neither
fragment suggests anything about what thinks amongst what is, fragment 3 gives being itself as the only possible agent of thought. Cornford’s subjectivist assumptions,
most apparent in fragment 8, disguise the panpsychist implication of fragment 3 and
transform it into a contest between a sanely epistemic account (only what is can be
thought) and an ‘hysterical’ Berkeleyan subjective idealist account (to be it is sufficient
to be thought).
By contrast, consider what Plato’s Parmenides puts in its titular philosopher’s
mouth: “all things think” (132c). Although the dialogue does not present this as a
statement of Parmenides’ own theory, but results from his criticism of Socrates’ presentation of the theory of forms, it effectively restates the panpsychist content of
fragment 3, but with a difference. While fragment 3 states that being is thinking, Plato’s
Parmenidean ventriloquism has it that every thing that is, thinks. While Cornford’s
accounts assume something about thinking (that thought presupposes a thinking subject), Plato’s assumes something about being (that it is composed of things). We can
express this contrast as between the subjectivist and the substantivist accounts.
Despite their contrast, the assumption that if there is to be thinking, there must be
a thing that thinks unites these accounts. From this assumption, there follows (a) the
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.3 (285)
Chapter 14. All things think
task of determining the nature of the “thinking thing,” with its well-known Cartesian
results; and (b) whatever its nature, its ‘thinghood’ determines it as bounded or finite.
Yet if we compare them with the fragment of which they are versions, we note that Parmenides makes no such assumption: we cannot, in fact, infer a thinking thing at all.
Rather than offering stipulations or hypotheses concerning the nature of what it is that
thinks, fragment 3 simply asserts the identity of the activities of thinking and being,
and specifies no subject or substance that either ‘can’ be or think, or in which such activities may inhere. Indeed, the substantivist assumption that Plato’s dialogue makes is
at odds with the essentially dynamic ontology advanced in the Sophist (247e): “I hold
that the definition of being is simply power.” Precisely this debate between Platonic
dynamics and an Aristotelian substantivism lies at the core of an important strand in
the contemporary metaphysics of nature. Recent discussions of powers have polarized
into grounded and ungrounded accounts, where the former hold that powers inhere
in substances, and the latter that they do not. For the latter, if nature consists solely in
powers, then since powers are not powers unless they can do something, that something that powers can do expresses a “physical intentionality.” With this conclusion,
however, there arises what one such theorist presents as “the threat of panpsychism”:
To extend the domain of intentionality from the admittedly mental sphere to what
are normally taken as purely physical states and properties, is to prove that Thales
may have been literally correct in attributing a soul to the magnet.
(Molnar 2003: 70)
What Molnar considers a “threat” is simply an argument: if there are no substances to
ground powers, then intentionality cannot be the property of a substance, but rather
the expression of a power. Hence intentionality is no longer the exclusive hallmark
of the mental, since it cannot a priori be said to inhere in any subject or substance
whatever. Regardless of the location of intentionality in all or some powers, we will
maintain this ungrounded dynamic understanding of nature in what follows.
For the moment, however, we have arrived at a core contrast between Parmenides’
identity thesis and the versions of it we have considered. The contrast is ontological
in nature, and requires the rethinking of both subjectivist assumptions on the part
of thinking, and substantivist assumptions on the part of being. Since we cannot infer
subjects or substances from the fragment, we are led to consider the kind of ontology
it actually proposes. And since in their infinitive forms, being and thinking are powers,
we may conclude that this is an ontology of powers, ungrounded in any substance in
which they might inhere or of which they might be properties.
Regardless, however, of whether the ontologies in question take substances/subjects
or powers as basic, Parmenides’ fragment leaves us with the further problem of determining whether ‘being = thinking’ amounts to one or many. Whether powers or
subject-substances, both dynamic and substantivist accounts assert a plurality of powers or substances. It is this problem that the subjectivist-substantivist seeks to resolve
by individuating the various existents and thinkers from the two otherwise undifferentiated infinitives. Yet subjects-substances are not the only ontological tools available
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.4 (286)
Iain Hamilton Grant
for this individuation. Reformulated in terms of powers, for example, the problem
remains: why, if being is identical to thinking, are there two powers between which
an identity needs to be forged (and what, apart from either being or thinking, does
Parmenides’ ontology make available to forge this bond)? Surely if the proposition
that being and thinking are the same is true, there are not many powers, but only
one? Either, that is, we are left with a monistic being-thinking which distinction must
ultimately be untenable, or with a dual-aspect monism of the Spinozist type.
This problem is therefore particularly pertinent as regards panpsychism, since
if being and thinking are the same, although the panpsychist claim is asserted, the
identity thesis does not so much resolve as restate the source of the problem. If a
panpsychist claim on this basis is to offer more, it must either assert Plato’s version
of the claim, assume the existence of particulars that think, and accept the task of determining their nature and extent (“all” or only some things?). Core to this version is
the “somatism” or substantivism the Eleatic Stranger uses the theory of powers to criticize in the Sophist. Or it may assert the identity thesis but with the qualification that
the identity expressed is itself dynamic rather than substantivist, concerning powers
rather than things, and thus expressing the necessity that being as such entails thinking. The cost of this latter account will be, as we shall see, that the asymmetry between
the two terms in the proposition is maintained – first being, then thinking – so that
being is not at all times ‘the same as’ thinking, but always entails that thinking ensues.
The conclusion that being and thinking are not always the same follows, indeed,
from both alternatives. Assuming that singulars think, the ‘pan’ or all in panpsychism
becomes unachievable in that it will not be the case that “being and thinking are the
same” unless “being” is reduced to “beings” and the “all” in panpsychism covers only
that collection of particulars, rather than being as such. On this account, the Platonic
“all things think” trumps the Parmenidean identity thesis, while demonstrating the
restricted ontological remit of the panpsychist claim. The restriction is informative on
two counts. Firstly, it is because of it that the problem of the nature of thinking particulars has become the focus of much panpsychist theorizing. Secondly, the ontological
perspective contextualizes and thus reorients the “either-panpsychism-or-emergence”
problem, and offers the conceptual space for a solution that can affirm panpsychism all
the way down without eliminating the genetic or natural-historical dimension emergence brings to the table. It is this latter that the dynamic understanding of the identity
thesis brings into focus, since it posits a temporal or genetic difference between being
and thinking. In what follows, we will work through both solutions, the finitist and
the dynamist. The prevalence of the former is due to what we might call the internalist
lure of philosophies of subjectivity and reflection. At its limit, this promotes a universal
phenomenology or subjective idealism that seeks to extend the structures of subjectivity to all that is thinkable, and to deny the existence of all that is not. The latter, by
contrast, enjoys the advantage over the former of providing a consistent ontology, not
least because it unites metaphysics with ‘physis’ or nature without denying a realism
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.5 (287)
Chapter 14. All things think
concerning ideas, and thus avoids the dualism we take it as the aim of panpsychism to
eliminate.2
If it is asked “why must panpsychism involve an address to nature?”, as if this prejudged any solution of the problem, the answer must be: if they are not nature, then
being and thinking – whatever their own relations – must be considered other than
nature, and a second level dualism ensues. If however the problem is considered from
the perspective of the question ‘does nature think?’, then its complexion changes. For if
nature does not think, dualism follows; if it does, then nature is capable of more than
the production of anoetic and inert substances with which it is usually and, in some
quarters, grudgingly accredited.
For now, however, we turn to the principle of finitude. The form in which finitude emerges as a problem for panpsychism is twofold: thinking substances (regardless
of whether these be minds or bodies) and reflective consciousness. The first binds
reason to particular (and therefore not to all) things, while the second imposes phenomenological or reflective access conditions on the identifiability of thought as such,
conditions to which only an actual (‘now occurring’) reflective consciousness has access. In both cases, a principle of finitude is used to derive ontological consequences
from the claim of insuperable epistemological limits, consequences that restrict the
plausibility of panpsychism a priori. To the complete contrary, we propose in what follows to situate the finitude of consciousness in a naturalistic ontology of powers rather
than substances that therefore supports both panpsychist claims and the temporal
anteriority of being to thinking that motivates emergentism.
. The principle of finitude
How can what forms the mere limitation of a science be made into the measure of
the groundedness of science in general?
(Schelling 1856, V: 137)
Schelling here poses the question we will pursue in this section. He poses it as a critical
question, affirming that this approach, while useful in the “subordinate sciences,” has
no place in philosophy, where “these limits do not exist.” Here, however, we will pursue
it as a programmatic rather than a critical question; we are looking, in other words, for
how this has been successfully achieved – how, that is, the Principle of Finitude has
been unquestioningly accepted by both post-Humean and post-Kantian philosophy.
. Although as part of his ‘broad church’ claim that panpsychism is less a theory than a
“meta-theory” of mindedness, Skrbina (2005: 2) asserts that a “panpsychist-dualism” is not
inherently contradictory on the grounds that a “Supreme Being” may have “granted a mind
to all things,” I will claim on the contrary that it is in fact contradictory, since it would simply
reconstitute dualism between original (Supreme Being) and derivative (all things) mindedness
without resolving the issue of the nature of thinking things. I am grateful to David Skrbina for
his insightful comments on an earlier draft of this paper.
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.6 (288)
Iain Hamilton Grant
The authors of the principle of finitude are Hume and Kant. Bluntly stated, this
principle states that possible experience provides grounds for the restriction of reason.
Neither Hume nor Kant deny that exceeding experience is a possibility for reason,
which first appears to Hume “unbounded by nature and reality,”3 but assert only that
in exceeding what can be accessed through possible experience, where experience is
defined in terms provided, in turn, by codifying its nature insofar as we have access to
it, reason loses all sure footing. The ‘possible’ in ‘possible experience,’ therefore, does
not so much promise rational access to possibilia as restrict reason to contingency, a
consequence we argue here follows from all ‘access’ arguments.
The key problem that the principle of finitude must address consists in deriving
necessary limits from contingent experience. Hume famously denies this is possible
at all, insisting instead on the “experimental method” of slow and patient trials of
experience. Since Kant, by contrast, seeks to bind the contingency of experience into a
framework of necessary laws, I will begin with him.
a.
Kant and the ‘I think’
Kant’s transcendental strategy consists in nothing more than the derivation of necessity from contingency. The fruit of this procedure is to divide the contingent content of
actual experience from the necessity of the formal laws of their constitution. The linchpin of these laws is the “transcendental unity of apperception.” “It must be possible for
the ‘I think’ to accompany all my representations [Vorstellungen 4],” Kant writes,4
for otherwise something would be represented in me which could not be thought
at all, and that is equivalent to saying that the representation would be impossible,
or at least would be nothing to me. That representation which can be given prior
to all thought is entitled intuition. All the manifold of intuition has, therefore, a
necessary relation to the ‘I think’ in the same subject in which this manifold is
found.
(1781–7, B131–132)
Kant here transforms my contingent lack of access to a representation into the impossibility of representation without access. It “would be nothing to me” therefore becomes
“it is impossible” because “I” is necessary. From this, of course, Kant proceeds to his
‘Deduction of the pure concepts of the understanding,’ which apply necessarily to all
my acts of sensible intuition. Yet Kant’s move is circular, for it asserts the necessity of
the ‘I think’ on the grounds of the impossibility of its non-occurrence, which is to say
the same thing: ‘it is necessary that x’ = ‘it is not possible that not-x.’ To avoid the accusation that this circularity is vicious, Kant adds a distinction between representation
that necessarily has no ‘I think,’ and a representation that contingently may not have
. “Nothing, on first view, may seem more unbounded than the thought of man, which not
only escapes all human power and authority, but is not even restrained within the limits of
nature and reality” (Enquiry ii, 13).
. Kant intends the term Vorstellung, translated as “representation,” to be construed as a
“placing-before” since this already includes that before which an empirical event is presented.
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.7 (289)
Chapter 14. All things think
one but will give rise to one. Attention is thereby shifted from a principle of finite access to a genetic account of the I. Kant calls this ‘transcendental unity of apperception’
a synthetic a priori truth, although being synthetic, i.e., produced, the product is made
to precede its own production. Disregarding this for the moment, the principle may
thus be reformulated: all representation necessarily gives rise to an ‘I think.’
Yet this may be taken in two ways, corresponding to its construal in the Critique
of Pure Reason and the Critique of Practical Reason, respectively. Firstly, it invites us
to check our representations for any absent cogito; when no such divergence is found
(since by it being my thought-experiment, it cannot be), we accept this as necessary
for our experience. Thus by virtue of this phenomenological quirk, the ‘I think’ is
smuggled into a synthetic truth that profoundly alters the concept of necessity. Hence
a ‘constant conjunction’ of thought with ‘I am thinking’ becomes associated with a
logical necessity attaching to possible experience. Only then are the pure concepts of
understanding deduced as the products and producers of a genetic account of I-hood.
Secondly, however, Kant gives no warrant to the extension of this phenomenological evidence to a necessary truth about us; his concern is famously with “finite rational
beings,” that need not be us at all, but will necessarily share with us those “pure concepts of the understanding” anchored by the necessarily accompanying ‘I think.’ Thus
the prospect of a warranted projection opens up that enables the increased speculative
range that Kant grants practical over theoretical reason. This projection seemingly also
lends support to the panpsychist cause, insofar as it does not in principle restrict cognition to one species of creature, but generalizes it to any and all entities capable of
generating an ‘I think.’ If we take Kant at his word, and consider the ‘I think’ a logical
necessity, then no entity whatever may be definitively ruled out as a candidate vehicle
for I-hood. It was left to Kant’s successor, Fichte, to generalize from this principle in his
‘Propositions for the Elucidation of the Essence of Animals’ (1800), where he notably
failed to extend this cognitive generosity to minerals.5
Although the second may seem, at first glance, to lend possible support to panpsychist claims, a feature uniting both of Kant’s accounts must be stressed: it is the
thinking I to which possible experience is restricted that warrants projection. Hence
the subjectivist trap into which projectivist panpsychisms fall: the barriers of possible
experience can be set wherever one chooses, on condition that the ‘I think’ be taken
as a necessary element for any entity to which ‘mindedness’ is thus extended. Hence
Sprigge’s recent (2006: 484, 478) claims on behalf of panpsychism:
It seems to me that it is only a panpsychist view of the world which can cope with
the two facts (1) that only experience exists and (2) that the physical world exists.
[But t]ry to imagine something which is unexperienced. Since physical things are
the most obvious candidates for things which can exist unexperienced, choose
some physical scene which is supposed not to be revealed to any mind. [. . .] It
. Fichte’s essay is found in Fichtes Werke, ed. Immanuel Hermann Fichte (1971, vol. XI: 362–
367). For a discussion of it, see Chapter 3 of my Philosophies of Nature after Schelling (2006).
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.8 (290)
Iain Hamilton Grant
seems to me evident that one cannot do so. [Thus] the physical world certainly
exists, but that it consists in innumerable interacting streams of experience.
Sprigge’s claims have the great virtue of making Kant’s procedure explicit: he enjoins
us to “imagine some physical scene. . . not revealed to any mind.” Since I cannot, the
world consists of streams of experience, of which imagining is one kind.
If this aspect of Sprigge’s claims highlights a procedural affinity with Kant, and
while both conclude that unexperienced reason is a priori impossible, their accounts
differ in one crucial respect. While Sprigge sets the limits of possible experience as
co-extensive with the world as such, Kant’s critical principles ensure that the only possible ‘non-I’d’ rational activity is as a moment in the process of the I’s emergence.
While leaving the extent of possible consciousnesses undetermined, therefore, Kant
limits consciousness to actually occurring consciousness insofar as this is the product
of a genetic process. This limited/unlimited relation will be important in what follows,
since despite himself, Kant’s derivation of necessity from contingency establishes the
finitude of consciousness with respect to an undetermined whole. In so doing, the
Principle of Finitude initiates a possible ground of discrimination between reason and
consciousness.
b. Hume and the Bounded Principle
Hume provides – and criticizes – one of the major arguments for which panpsychism
is routinely attacked by its detractors. Sometimes called the argument from analogy,
we shall call it the mereological or extension argument, since it concerns the legitimacy
of extending what may be true of the part to the whole. While in the Dialogues this
preoccupies Philo as he discusses the determinability of the origin of the universe, our
present focus concerns the determinability of thought: “What peculiar privilege has
this little agitation of the brain which we call thought, that we must thus make it the
model of the whole universe?” (Hume 1779/1993: 50).
For Hume, the point may be crucial insofar as it seems to crown the naturalism
of the Treatise with a proto-neurological account of thought itself. In Philo’s brief
statement there occurs a shattering in the manifest image of thought that dismisses
another anti-panpsychist argument: the argument from access. For here Hume’s naturalism achieves a richness in consequence that belies the brevity of its exposition.
Consider the following passage in which Philo is building his case against Cleanthes’
design argument:
Thought, design, intelligence, such as we discover in men and other animals, is no
more than one of the springs and principles in the universe, as well as heat or cold,
attraction or repulsion, and a hundred others, which fall under daily observation.
(ibid.: 49; my emphasis)
It is so quick as almost to escape notice: Hume considers thought as “one of the springs
and principles in the universe”; in other words, as a natural power that, as with all
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.9 (291)
Chapter 14. All things think
the other powers in nature, must remain as epistemically indeterminable as they are
ontologically undeniable.6
Its naturalism notwithstanding, Hume’s project has evident parallels with the
transcendental strategy Kant proposed: to discover the laws of thought, thinking must
take thought as its object. Indeed, in the immediate wake of Kant’s critical revolution,
its most vociferous avatars were naturalists: Christoph Girtanner in natural history,
and Johann Christian Reil in psycho-physiology.7 The latter in particular considered it
essential to completing Kant’s program that the many powers of mind be reduced to a
single one, the same as operates throughout nature.
While Kant’s successors share Hume’s naturalism concerning thought as one of
nature’s “springs and principles,” it is in practice less its nature but the sources of its
content that forms the basis for Hume’s application, in the Treatise, of the Baconian
“experimental” method in natural philosophy to its moral counterpart. Rather than
constituting a problem of insufficient sampling as it would in other experimental contexts, this procedure is unavoidable insofar as the “exact analysis” of the “powers and
capacity. . .of the human understanding” (Enquiries I, 7) leaves us no other possible
basis than reflection upon its operations; and we have no other point of reflective access or observation but our own thinking. Thus the nature of thought, insofar as this
is determinable at all, is to be determined in accordance with the character of our
thinking.
However much, then, Philo may protest against Cleanthes taking “the operations
of one part of nature upon another for the foundation of our judgment concerning
the origin of the whole” (Hume 1779/1993: 49), the principles of the experimental philosophy entail that Hume must persist in so doing. Indeed, he confirms that analogy is
the foundation of “all our reasonings concerning matter of fact” (Enquiries I, 9) while
acknowledging it a “weak [and] bounded principle” (1779/1993: 50). In effect, Hume’s
Bounded Principle asserts rather than denies the mereological argument – extending
the operations of one part of nature on another to the whole – as made necessary by
the contingent fact of limited access.
Taken together, the Bounded Principle and the Finitude of Consciousness form
the Principle of Finitude, which therefore asserts:
1. that conscious is finite with respect to reason, and
2. that this finitude is necessary
. It is this claim that lies at the core of a renewed interest in Hume as metaphysician. See R.
Read and K. Richman, eds., The New Hume Debate, 2nd edition, revised (2007).
. Christoph Girtanner was the author of On the Kantian Principle for Natural History (1796),
and Johann Christian Reil’s power-monism is expressed in his On the Vital Force (1796). For
more, see my ‘Physics of analogy,’ in R. Jones and A. Rehberg, eds, The Matter of Critique:
Readings in Kant’s Philosophy. (2007: 37–60).
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.10 (292)
Iain Hamilton Grant
Unlike Kant, Hume’s Bounded Principle does not seek to derive necessity from contingency, but asserts contingency as necessary. Whichever way round, however, this
equation is the chief achievement of the Principle of Finitude. At issue between them,
for present purposes, is whether the finitude of consciousness vitiates an account of
panpsychism. Before we turn to this, we will examine a more recent contributor to
debates surrounding the Principle of Finitude.
c.
After Finitude?8
In a recent work, Quentin Meillassoux (2008a) has asserted precisely the necessity of
contingency in support of a renewed “speculative” approach to the problems of ontology. Like Hume and Kant, Meillassoux considers metaphysics a catalogue of rationally
insoluble problems. Unlike either Hume or Kant, he wishes to end the imposition of
finitude projected from “thought’s discovery of its own intrinsic limits” onto reason itself, a projection whose dominance he considers the principal achievement of modern
philosophy. “Whereas the Parmenidean postulate, ‘being and thinking are the same’,
remained the prescription for all philosophy up to and including Kant,” Meillassoux
writes (2008a: 44), “the fundamental postulate” of more recent, ‘post-metaphysical’
philosophy has been: “being and thinking must be thought as capable of being wholly
other.” In other words, metaphysics is no longer credible precisely because there is no
necessary relation between thinking and being. This being so, it can and should be
abandoned, he recommends, by any philosophy interested in reason and reality, that
is, a speculative philosophy that begins not from the problem, but from the fact of
existence.
Speculative philosophy must therefore abandon the search for a sufficient reason
“why what is, is the way it is” (2008a: 82) since no such reason is to be found. This
being so, the “principle of sufficient reason,” which states that there must be a reason
for the existence and nature of everything that is, can only be an object of belief, making metaphysics into “fideism.” The unavailability of such a reason entails not only
that no reason can be supplied for the way things are, but also that the way things
are is a contingent matter: they may be otherwise, or change tomorrow, or haphazardly, without warning. It follows that “contingency alone is necessary” (2008a: 65).
That is, it is necessary that there is no necessity attaching to the laws of nature remaining the same. While there may be causes that in fact form the laws of nature,
Meillassoux agrees with Hume that the nature of the problem is the “rational justification of our belief in natural necessity” (2008b: 272; my emphasis), although as has
been recently remarked (Strawson 1992), Hume’s metaphysics contains a commitment
to the indemonstrable existence of causal powers, so that where “reason is incapable. . .,
nature herself suffices” (Treatise I.iv.7); Meillassoux’s, as we shall see, does not. Finally,
rather than resting within the insuperability of the possible difference between being and
thinking, Meillassoux advocates, in an extended analysis of ‘Hume’s problem,’ that we
. I am grateful to Jeremy Dunham for his stimulus in this direction. See his ‘Quentin Meillassoux and the End of Metaphysics’ (forthcoming).
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.11 (293)
Chapter 14. All things think
transform our perspective on [the resulting] unreason, stop considering it as the
form of our deficient grasp of the world and turn it into the veridical content of
this world as such – we must project unreason into things themselves. (2008a: 82)
If Meillassoux’s abandonment of sufficient reason is as absolute as he here claims, however, why is it additionally necessary to “project unreason into things themselves?” For
surely if no necessity governs “why things are what they are,” then “unreason” or, less
dramatically, the non-existence of a sufficient reason lies not merely in the “deficiencies of our thinking” that Hume analyzed, then where else can it lie but in the things
themselves? Yet if unreason did indeed lie in the things themselves, then wouldn’t this
unreason itself be ‘sufficient’ in precisely the manner that the principle of sufficient
reason stipulates? Meillassoux’s dilemma, in other words, is the following: either unreason lies in things, in which case unreason is sufficient in the metaphysical sense; or
it does not, in which case it lies within the deficiencies of our thinking. If the former,
speculative reason is a species of metaphysics rather than an alternative to it; if the
latter, it has not escaped the Principle of Finitude.
The point is important because Meillassoux has returned our attention to the relation between thinking and being as lying at the core of our problem. He goes so far
as to define post-Kantian (and, by extension, post-Humean) philosophy in general as
premised on the possibility that thinking and being are entirely unrelated. However, in
the end, the projection argument belies the principle of finitude operative in his solution to the problem of sufficient reason: reason is, in Hume’s words, not amongst the
springs and principles of nature.
While this may seem at odds with After Finitude’s opening Cartesian reconstruction of Locke’s distinction of primary and secondary qualities, it in fact illustrates
an important consequence of this mathematical defense. Seeking a new, non-Kantian
way to repair the gulf between rationalism and empiricism, Meillassoux presents what
“can be formulated in mathematical terms” as “properties of the object in itself ”
(2008a: 3), and defends these against two alternatives. Firstly, any ontological attachment to subject-dependent sensible properties (qualia); but secondly against things
independent of either mathematical formalization or perception. This is clear if we
compare Meillassoux’s distinction between mathematizable and sensible properties
with its Lockean source. Locke calls primary qualities those “resemblances” of the
qualities of bodies “produced in us” by non-resembling secondary qualities, and adds
“There is nothing like our ideas existing in the bodies themselves” (Essay II.viii, 16).9
Thus the purpose of the distinction for both Locke and Meillassoux, is to divide reason
from nature. The difference is, in the process of isolating ideation from nature, Locke
acknowledges the relation between non-resembling bodies and ideas as a genetic one;
Meillassoux, by contrast, makes the division absolute and unbridgeable.
. Hume agrees, concluding what Molnar (2003: 114–115) describes as his “projectivist” account of causality, thus: “Upon the whole, necessity is something that exists in the mind, not in
objects” (Treatise I.iii.14).
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.12 (294)
Iain Hamilton Grant
In this sense, all philosophies beginning from the finitude of reason turn into questions concerning cognitive, experiential or imaginative access to a nature necessarily
external to it. It is therefore assumed from the outset that thinking has no necessary
relation to nature. Yet this is false unless we consider thought to arise outside nature.
If we do not accept this, but affirm thinking as consisting of events in a physical world,
some version of panpsychism becomes inevitable. It is to this we now turn.
. Panpsychism, finitude and externalism
In the foregoing section, we have seen that transcendental philosophies (the Kantian
tradition) and experimental philosophies (the Humean tradition) that take consciousness or exclusively human reason as their starting point, have as a consequence of their
principled finitude an entailment of externality for which they cannot account. The
principle of finitude, as we have constructed it, is an excellent example of what Molnar
calls “inversion hysteria.” Citing as examples of this condition Mill’s inversion of “it is
possible to perceive real objects” into “objects are a permanent possibility of sensation”
and Quine’s insistence that the “conceptual centrality of logic and maths to science is
all there is to necessity,” Molnar describes this process in the following terms:
Some piece of objective reality has characteristic effects on and in humans. You
then turn around and define this piece of reality in terms of its effects on humans,
thereby making it mind-dependent. Inversion hysteria is a kind of subjectivizing
of reality, a kind of subjective idealism. (2003: 223)
Accordingly, what we contest here is not the finitude of consciousness, but rather that
it can be derived from consciousness itself. If it cannot, then its conclusions cannot be
asserted necessary for reason.
This is why, in the first section, we followed the Parmenidean problem concerning thinking and being as such, rather than under the constraint of the Principle of
Finitude. Of the two metaphysical trajectories that follow from the identity thesis,
the substantivist one leads towards either a panpsychism premised on a plurality of
thinking things, or to a finitism premised on the one thinking thing to which we have
access. The other, dynamist trajectory leads either to a compatibilism regarding the
two powers under discussion (the power to exist and to think), or to a naturalism that
has thinking as a genetic entailment of being. In this section, we will attempt a portrait of this latter, and provide a reason for it other than the now obvious limitations
of finitism.
We begin with a quotation from the British Idealist philosopher, Bernard
Bosanquet.
Pan-psychism seems to me a gratuitous hypothesis, depending on a hasty resolution of the responsiveness of nature to mind by help of the idea of resemblance,
and wholly failing to recognize the complementary functions of subjective mind
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.13 (295)
Chapter 14. All things think
on the one hand and externality on the other as together essential to any complete
form of conscious experience.
(1912a: 364–365)
There are two reasons we might offer as to why Bosanquet considers panpsychism “a
gratuitous hypothesis.” The first of these is that his attention is focused here on the
requirements of an account of the “complete form of conscious experience” which,
he claims, must involve “externality,” which we will address below. The second is that
panpsychism seems, in precisely the manner we noted concerning Parmenides’ identity
thesis, merely to affirm a changeless identity of the “subjective mind” and “external”
sides of experience, while explaining neither. The question is, why does this identity
not satisfy the “complementarity” criterion Bosanquet invokes?
The answer becomes apparent when we consider Bosanquet’s concept of externality, by which he refers to nature qualified by temporality. Externality thus designates
“the first nature of all” (1912b: 84), that which “comes first” (1912a: 219). Externality is
“the instrument for sculpturing minds” (1912b: 16), and is “the source and storehouse
of all primitive properties, contents, and distinctions of mind” (1912a: 359).
Two questions arise regarding the thesis of externality. The first is how it differs,
if at all, from the standard emergentist rebuttal to panpsychism? The second is more
complex, as it involves an address to the problem of natural causality, and therefore
concerns the rational warrant for the sequential dimension of externality.
Taking the questions in order, emergentism standardly claims that mind is not
present all the way down, but emerges as a consequence of sufficient complexity being
attained in the physical architecture necessary to support mindedness. The problem
here concerns the causal triggers responsible (a) for complexification and (b) for
mindedness at all, since it is not present beneath a given complexity threshold. Emergentism explains mindedness, therefore, only if it can explain these causal triggers.
Recent emergentisms have addressed this problem through the concept of ‘autopoiesis’
or ‘self-organisation.’
Conceptually, however, this concept assumes what is to be explained, since the
‘self ’ of ‘self-organisation’ does not precede but consists in its emergence from the
accretion of processual richness. The emergentist’s point in fact concerns the unpredictability of the emergence of these self-organizing processes, which is why many
affirm the reintegration of physics and history, or the renewed primacy of ‘becoming’
over ‘being’ (Prigogine & Stengers 1984). Since the consequence of this argumentation is that the explanation of sufficient complexification is sacrificed, effectively, to
an after-the-event account – a history – of particular emergences, the emergentist
cannot explain the causal triggers involved in emergence, but affirms instead the unpredictability of natural history.10 Since to assert ‘unpredictability,’ however, does not
. Matters are more complicated, however. Consider, for instance, Isabelle Stengers’
(1997: 22–23) account of the distinction between the “phenomenological” and “fundamental”
versions of nonlinear processes: “When I learned physics, I accepted as ‘only phenomenological’
the laws that describe ‘irreversible’ evolutions – that is, evolutions that take place in one direc-
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.14 (296)
Iain Hamilton Grant
provide criteria for complexification, but reaffirms only the want of an explanation, the
emergentist has no rational warrant for maintaining that there is a threshold beneath
which mind cannot arise.
So if the genetic dimension of the externality claim is a problem for the panpsychist, it remains one for the emergentist, who enjoys no particular explanatory advantage over her rival. It is to this problem we now turn. The externality claim, as
we have seen, does not simply involve an objectivity complementing a subjectivity,
but rather the priority of nature. Priority and posteriority establish asymmetrical relations between events which invokes a causal, or genetic processes. Bosanquet’s further
claim that nature is the “storehouse” not only of the production of mind, but also of
its “contents and distinctions,” makes the externalist claim more complex in that it
holds out the prospect of a causal map of each act of intellection, from geological to
neurological events.
The genetic account is only one part of the externalist thesis; the complementarity
criterion asserts that posteriors (whatever their nature) necessarily have this externality, and that it is insuperable (Bosanquet considers the panpsychist hypothesis “hasty”
precisely because it is insensible to priority and posteriority), since if they did not,
there would be no prior-externality. Consider, for example, the relation between neurological processes and actual ideation. The non-resemblance identity we established
via Parmenides holds good here, and since only an ‘hysterical’ finitist would hold to
the ontological separability of the ideational qualia from the neurochemical activities
that produce them, it clearly illuminates both dimensions of the externalist claim: logical and genetic priority coupled with an externality between the priors and posteriors
thus related. Further, since Bosanquet claims nature to be the source and storehouse of
mind, we can see that the externality thesis does not entail any ontological separability of mind from nature, and that it posits only priority as its threshold criterion for
“mindedness.” In other words, it affirms the identity thesis dynamically understood,
as we outlined earlier in this chapter, and explicitly denies that resemblance plays any
role in identity relations. Having examined the finitist position, we are now in a position to see why this is the case: resemblance relations are necessarily phenomenal,
and entail therefore a second-order posterior, or a third element, to formulate them: a
consciousness.
Prior to consciousness, the identity relation already involves nature (necessarily)
and “the contents and distinctions” of mind. In other words, nature is prior and external to reason, which is in turn prior and external to consciousness. Externality is
satisfied in every relation, so that reason can be said to be less ‘sparse’ than consciousness. Hence the advantage of externalism over finitism: the latter requires to
tion (a mixture does not ‘unmix’ itself. . .). From a fundamental point of view, the differences
between the evolutions that we observe and those that we think impossible is not valid. . .. The
fundamental laws of physics do not recognize what leads us to recognize without hesitation that
from one situation another will follow. It gives no direction to what we traditionally call the
‘arrow of time’.”
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.15 (297)
Chapter 14. All things think
restrict reason to the “bounded principle” of consciousness, which can deal only in
resemblances, projections and analogies. The advantages of externalism for panpsychism are equally evident: nature necessitates reason and consciousness and indeed,
on condition of these genetic externalities, is identical to them.
We now move on to the genetic or causal claim on the part of externalism. Hume
clearly demonstrated that, as long as we remain within the bounds set by the experimental philosophy, causality remains an hypothesis grounded only in an anthroponoetic habit, and therefore not grounded at all. Grounding causality, therefore, will
involve the abandonment of that philosophy. We have already encountered an alternative version of the causal thesis, one bound up with the principle of sufficient reason.
Leibniz formulated the principle of sufficient reason in a number of ways, of which
we will cite two (thus ignoring its application to God’s selection amongst possible
worlds). In one form, it states that “there can be no fact real or existing, no statement true, unless there be a sufficient reason why it should be so and not otherwise”
(Monadology, section 32). So stated, it involves the relation between reason and existence, although it does not specify the nature of this relation. In another form, however,
it states “the full cause is equivalent to the entire effect” (cited Stengers 1997: 25). In
this latter version, it specifies the nature of the relation as equivalence. How are we to
understand the assertion of the equivalence of reason and existence, or of causes and
effects? Clearly they differ, as there is more in reason that in existence (there are rationally possible non-existents), and more causes than have actual effects (or causes
would no longer be causes). Moreover, if these are not two different principles, but
one and the same, then the principle of sufficient reason involves equivalence relations
between reason and existence, and cause and effect. Accordingly, to deny the principle
would be to deny that the reason for existence is a causal one, and that existence is
always the effect of reason. It does not stipulate which reason, nor which cause, merely
that for a reason to exist is for causes to involve effects.
We propose that the externality thesis satisfies reason in that it makes nature equivalent to reason, on condition that equivalence is dynamically understood in the way
we have outlined it. Once it is so understood, this equivalence is clearly causal. In Parmenidean terms, being = thinking entails that being generates thinking; in naturalistic
terms, nature = reason entails that nature generates reasoning.
How, finally, does this leave the panpsychist claim? Having addressed the genetic
and the complementarity dimensions of the externality thesis, we may note the following. Firstly, by the genetic dimension, nature thinks, or reasoning is one of the powers
causally operative in nature. Secondly, by the complementarity dimension, thinking
is finite with respect to the nature that generates it. Thirdly, if (nature = reason) =
consciousness, then consciousness cannot account for its own genesis (Hegel’s error
and Schelling’s insight), which can only be explained by a reason that exceeds consciousness (is not access-dependent) and is exceeded in turn by a first, a nature, that
generates it and of which it forms part.
Before concluding, we return to Meillassoux’s arguments where we left them. His
claim is that since the principle of sufficient reason cannot be satisfied, unsatisfied
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.16 (298)
Iain Hamilton Grant
reason must be turned into an ontological principle. In other words, the reason of
existence is ungrounded. We do not wish to dispute the ungrounding thesis, but simply
to assert that this ungroundedness is nature.
Core to contemporary debates concerning the metaphysics of nature is whether
powers are regarded as properties of entities, or are ontologically basic. This is known
as the problem of ground, otherwise known as the problem of sufficient reason. Introducing a physicalist understanding of this problem that was absent from, although
implicit in, our discussion of Meillassoux’s treatment of it, Molnar writes:
The ground of a power, P, is the set of properties (all of which are conceptually
distinct from P) by virtue of which a thing has P. The Thesis of Groundedness is
the claim that necessarily all powers have grounds. I have argued. . . that this claim
is falsified by the basic powers of the fundamental subatomic particles that appear
to be ungrounded or pure dispositions. (2003: 147)
Although Molnar goes on to express doubts concerning the uniform applicability of
the ungrounded argument to powers in general, the cost of its rejection is a dualism
of property and entity that in effect denies powers their power-ness. This is because
to be a power entails that it be a power for something, a disposition towards something that must, by virtue of being a power at all, be (at least) possibly operative. The
cost of ungroundedness is a “physical intentionality” which entails, so Molnar, the
‘threateningly panpsychist’ conclusion that Thales was correct, and magnets are indeed “ensouled” (1993: 70). The dualism of power and substance is doubly apparent
in the hypothesis of the ‘magnet + soul’ complex. Firstly, a powers ontology problematizes the spatiotemporal localization that makes an assertion of the type ‘this is a
magnet’ possible. Rather than enabling an a priori distribution of discrete substances,
that there is a magnet presupposes a solution to this problem in a nature composed
of powers that Molnar has not provided. Secondly, that nature is composed of powers
does not entail the existence of souls at all, but only a qualitatively variable continuum
of physical intentionality, amongst which magnetism and thinking must number.
This second dualism, therefore, is apparent in the division of thinking from other
physical intentionalities. As with the first dualism, a powers ontology provides no
grounds for so doing. If, by contrast, powers are ungrounded, physical intentionality
confirms even as it naturalizes Parmenides’ identity thesis: when there is being, then
there is thinking. Molnar’s “threat” should therefore have concluded: ‘when there is
magnetism, thinking must follow.’ Thus, by rejecting the dualism of reason and the
other powers constituting nature, genetic externalism yields an ontology that confirms the principle of sufficient reason in each of its genetic stages, while entailing
ungroundedness as a consequence of their nature.
TSL[v.20020404] Prn:5/11/2008; 10:40
F: AICR7514.tex / p.17 (299)
Chapter 14. All things think
. Conclusion: Speculative physics and the ensouled magnet
We have argued that the genetic and the complementarity dimensions of externalism,
which entail a dynamic ontology, accounts for the real existence of finitude and for a
panpsychism all the way down, that is, without exception. This is because a rigorous
dynamism entails the derivative nature of individuation, since individuation, or the
production of products, is necessarily consequent upon production rather than prior
to it. However, dynamism, although ‘all the way down’ never reaches ground, since
were it to do so, powers would cease to have the ontological priority over their products, and would become instead modes or properties of a substance or subject whose
nature would have to be determined all over again. That the ‘pan’ of panpsychism may
have to be rescinded, therefore, by virtue of the genetic element of the externality thesis, this sacrifice recovers panpsychist claims from the anti-naturalist assumption that
being = thinking constitutes an extra-physical relation, and thus makes conceptual
room not for emergence as regards ‘mindedness,’ but for the accommodation of a naturalism that insists on the priority of physical production over product. We may thus
take the naturalist sting out of the emergentist tail.
We have also argued that the problem of ground is most completely addressed
through the thesis of ungrounding, both as regards the powers of nature and those
of reason. This is because an analysis of ground that concentrates only on its rational
dimension (the principle of sufficient reason) to the exclusion of its causal one ‘denaturalizes’ reason and cannot answer, therefore, the problem of its own generation.
Regarding, therefore, what many will consider an implicit dual-aspect theory involved
in the assertion of two classes of powers, the natural and the rational, we have argued that this is soluble by the genetic externalism we have drawn from Bosanquet,
on the one hand, and the powers theorists, on the other. This does not mean therefore that reason must simply be naturalized in the manner to which much twentieth
century philosophy became accustomed, but that the identity relation between being and thinking be conceived genetically. Geneticism entails in turn that the ground
of thinking is being, and that of being is power, as the Eleatic Stranger proposed. A
powers ontology, however, consistently carried through, entails that the powers are
ungrounded. It is this that enables the panpsychist to eat physical rather than merely
noetic cake.
Finally, by avoiding the internalist lure integral to the phenomenologizing, epistemologizing, or multiple minding solutions to the panpsychist problem, with their
entailed subject-substantivism, reason is no longer tied to a particular class of entity.
Whereas the finitist can only analogize or project intelligence onto plausible vehicles for a rationality judged by resemblance, externalism makes the following case
plausible. When we say nature causes thinking, the thinking in question is no longer
localizable within any given subject-substance. This being so, the causing of thinking
is not as sparse in nature as the finitist weakly imagines.