The Powers Metaphysic Neil E.

Williams
Visit to download the full and correct content document:
https://ebookmass.com/product/the-powers-metaphysic-neil-e-williams/
The Powers Metaphysic
To Charlie, for introducing me to powers,
and to Kelly, for helping me exercise them.
Acknowledgements

A great many people have helped me in the writing of this book. So many that there is
no way I can thank them all (sorry!), but I want to briefly thank a few who have really
helped this book come to fruition. A big thanks must be extended to C. B. Martin, the
great patriarch of the powers theorist family, and my undergraduate teacher some
two decades ago. It was Martin who first introduced me to metaphysics, to disposi-
tions, and to powers. I owe a great deal to Martin, not just in terms of my thoughts
about powers, but for instilling in me an ontology-first approach to metaphysics and
philosophy in general. Some of the ideas in the book grew out of acorns planted in
my dissertation at Columbia University. I was most fortunate to have the guidance of
a committee that included Achille Varzi, John Collins, David Armstrong, and Peter
Unger. More recently I have benefited from numerous discussions with my col-
leagues at Buffalo, and from the many graduate students on whom most of this
material was first tested. Three of those graduate students deserve special mention for
having worked through more than one draft of the book and for providing numerous
comments on its content and its writing. They are John Beverley, David Limbaugh,
and Robert Kelly. Others who have read full drafts of the book and helped it along the
way include: Matthew Slater, Chris Haufe, Michael Brent, Tyler Hildebrand, Travis
Dumsday, and Carl Gillet. I also want to thank the anonymous referees from
OUP. Their comments were thoughtful and extensive. Lastly, there are lots of people
in lots of places who I have made listen to parts of the book over the many years at
conferences, seminars, workshops, and over cocktails, and who have provided all
sorts of feedback. To them, and to anyone I may have missed, I offer a large collective
thanks. Your help is truly appreciated.
The epigraph at the beginning of Chapter 8 is reprinted from Theodore Sider, Four-
Dimensionalism: An Ontology of Persistence and Time. Copyright © 2001, Clarendon
Press. It is reproduced by permission of Oxford University Press through PLSclear:
https://global.oup.com/academic/product/four-dimensionalism-9780199244430.
 1
Revolutionary Metaphysics

1.1 The Soft Sell

THE SOFT SELL
-: Suited SALESPERSON knocks on the door of an average suburban
home. Slightly dishevelled HOMEOWNER opens the door and stands
impatiently in the doorway.
: Good evening! Do you mind my asking if you are happy with
your vacuum cleaner?
: I guess so . . . it gets the job done. (pauses) Hold on—are you
selling something? Now’s really not a good time for me to . . .
: (interrupting) No! Not now anyway. I’m in the sales game, but
that’s not why I’m here today.
: Then how can I help you?
: I’m here to talk to you about your vacuum. You say you are
happy with it; do you recall when you got it, or why you chose it?
: I got it from my parents some years back; it’s the newer
version of the one they have. I’m happy enough with it, but I wouldn’t say
I chose it. It was the one I’d grown up with, plus a few updates, mostly
cosmetic. I guess I didn’t even know there were other vacuums. But mine’s
fine thanks.
: steps back and starts to close the door. SALESPERSON steps
forward.
: (hurriedly) Most people don’t know they have a choice. Most
people, like yourself, use the same vacuum they grew up with. But there’s
another vacuum on the market; I’m here to let you know that.
: Like I said, I’m pretty happy with . . .
: (continuing) And this vacuum doesn’t use bags—it’s bagless.
No expense or annoyance of buying new bags; no messing with removing
filthy full ones. And it has attachments for stairs, and for pet grooming, for
ceiling fans, for cleaning up spills, and plenty more, too.
: Bagless? Hmmm. That’s something I could go for; I hate
changing the bags. And an attachment for cleaning ceiling fans? My vacuum
can’t do that.
: Then you’ll also be interested to know that this new vacuum
outperforms yours on a range of tasks.
: Sounds like this new vacuum is better than mine!

The Powers Metaphysic. Neil E. Williams, Oxford University Press (2019). © Neil E. Williams.
DOI: 10.1093/oso/9780198833574.001.0001
  

: I think it is, but I won’t deny that the old one does a good
job and has some attractive features. I just think the new one is prefer-
able overall. I have one with me in the car: would you like to take a look
at it?
: I’d mostly just be interested in the ceiling-fan attachment. Can
I see just that?
: Unfortunately . . .
: (interrupting) Package deal, eh? Gotta buy the lot to get the
small part I want?
: It’s not that I wouldn’t show you just the ceiling-fan attach-
ment if I could, but it’s not compatible with your vacuum. I’d love to be
able to take the best of your vacuum and combine it with the best of this
new one, but there’s simply no à la carte option: the two systems aren’t
alike.
: So it’s one or the other?
: Yes. Just like yours, there are a few optional components, but
there’s no mixing and matching across brands. I take the new one to be
superior, on balance. But don’t take my word for it. Why don’t I leave it with
you for the weekend, and you can decide for yourself? See how it compares
to your current one; I think you’ll be surprised. There are some pretty neat
features on the new one, and you might find that you aren’t as happy with
your current one as you think. Then again, you might decide that you are
happy enough with what you’ve got, but at least you’ve had the chance to
think about it and make a choice.
: Sure, I’m game.
Unless you are reading this first chapter out of context, you will know that this is not
a book about vacuums, and that I am not in the business of selling them. But I do
have something to sell, and it, too, will be a soft sell. This is a book about metaphysics,
and what I am selling is a metaphysic.
A metaphysic (note the lack of terminal ‘s’) is what you get when you embed a
fundamental ontology within a larger metaphysical framework by repeatedly
appealing to elements of that ontology in explaining metaphysical phenomena. At
its most basic, an ontology is an answer to the question ‘what is there?’ To respond,
all one need provide is an inventory of the types or categories of things that make
up fundamental reality. It is a laundry list of the types of entities countenanced in
the most basic theory of the world, along with a description of their natures.
A metaphysic, however, does a great deal more. A metaphysic explains how the
items on the list are connected to one another, and tells the story of how the items
on the list work together to produce the world around us. An ontology is a list of
ingredients, but creating a dish requires a recipe that tells us how those ingredients
are to be treated, what measures they come in, what directions one must follow to
combine them correctly, and what their assemblage will produce. It is only when
we see an ontology in the context of a grander theory that we can begin to see what
it is worth.
    

One such metaphysic dominates not just contemporary metaphysics, but

contemporary philosophy as a whole.¹ This dominant metaphysic is called the
‘neo-Humean metaphysic’ (or ‘neo-Humeanism’), owing to its Humean ancestry.
According to it, properties are inert, causation is illusory, and possibility is
unbounded. The neo-Humean metaphysic was originally adopted for what then
seemed like adequate reasons—such as its compatibility with a strong form of
empiricism in vogue in the early modern era—but many of those reasons have
long since been forgotten, and others no longer apply. In the period that followed,
neo-Humeanism became the status quo, and remains so. This was due, in no small
part, to its success at dealing with problems. But its establishment was further
aided by the lack of any serious competitor. This latter reason is far from ideal as
justification: a metaphysic is supposed to capture the fundamental aspects of
nature; it is a posit about the world at its most basic. Thus, our confidence that
a metaphysic gets things right goes up when it is adopted critically, selected as best
from among viable alternatives. There has not been an alternative for quite some
time. This makes it harder to be critical, meaning that we tend to overlook or
downplay flaws, perhaps even embracing them as features. After all, how can you
treat aspects of a metaphysic negatively if you are unable to recognize that
metaphysic as replaceable? Without a viable alternative it gets harder to appreciate
that the neo-Humean metaphysic is a posit at all.
Having an alternative, of course, is not to suggest that which metaphysic is right is
in any way up to us. We also choose among scientific theories, and that no more
justifies attitudes of anti-realism or scepticism in science than it does in metaphysics.
But gaining that critical eye, and having a real choice, demand that we have a credible
alternative. Up until recently, neo-Humeanism was the only game in town. But in the
last few years a serious competitor has started to take form. We do have a choice.
That alternative comes in the form of the powers metaphysic.² The powers
metaphysic is built on a radically distinct ontology from that which props up the
neo-Humean metaphysic, replacing the neo-Humean’s impotent properties with
inherently powerful ones. These powerful properties—powers—bear strong relations
to other properties, and are genuine causes of their effects.³ From the powers
ontology emerges a very different picture of the world. Unlike neo-Humeanism,
the powers metaphysic paints the world as dynamic: it treats action as coming from
within; it puts the oomph back in causes; and treats possibility as genuine and

¹ Or so I contend. Informal polling among non-metaphysicians finds that folks often take for granted
many of the theses discussed in Chapter 2. Groff (2012) argues that a good deal of contemporary social and
political philosophy is built on a neo-Humean ontology. And though many metaphysicians reject the views
they credit Hume with holding, the same does not apply to neo-Humeanism.
² The already classic works in the field include, but are not limited to: Harré and Madden (1975),
Shoemaker (1980), Mumford (1998), Ellis (2001), Heil (2003), Molnar (2003), Bird (2007a), and Martin
(2008).
³ My powers are what sometimes get called ‘dispositions’. I prefer the use of ‘power’ over ‘disposition’
because the latter has been less carefully used. However, as so much of the relevant literature speaks of
‘dispositions’, I am forced to use that term at times. We will see in Chapter 3 that I have a more refined way
of separating them.
  

bounded. As we shall see, it is a strikingly different metaphysic to that which

currently holds sway.
I am a friend of powers and an early adopter of the powers metaphysic. Like neo-
Humeanism, it has its costs and its warts, but I believe that on balance it comes out
ahead. I cannot help but think that others ought to join me in choosing the powers
metaphysic over neo-Humeanism, but getting you to join me is not the primary
aim of this book. My aim here is the soft sell: I want to make clear that there is a
coherent alternative to neo-Humeanism. I want to tell you about the alternative,
have you think about it, and have you come to see the neo-Humean metaphysic as
replaceable. And I want you to cast a critical eye over both. That way your choice is
genuine. And though at times I indicate aspects of the neo-Humean metaphysic
that look more like flaws than features, it is not my plan to construct devastating
objections against it. In fact, we should be sceptical of anything presented as a
devastating objection against neo-Humeanism as a whole: it is a solid, coherent, and
all-round attractive metaphysic.⁴
At this point one might wonder why having an alternative leaves us with a choice,
rather than an opportunity for integration. Why not take the best of each? The
answer is that central parts of a metaphysic connect to one another in such a way that
mixing is difficult. As detailed in §10.4, the parts rely on and inform each other in
such a way that they are largely package deals, much like the vacuums above. There
are options, but there is little room for combination. Such is the nature of systematic
metaphysics, as I will soon explain.
I believe that on balance the powers metaphysic is to be preferred, but that is
something you will have to come to on your own. Or not—maybe after seeing the
alternative you will stick with the status quo. That is fine, too, as long as the result is a
more critical adoption of the dominant metaphysic. My job is to flesh out the powers
metaphysic and the ontology at its core so you can appreciate its worth. I will then
leave it with you for the weekend to try out.

1.2 Systematic Metaphysics

I approach the question of which metaphysic the reader ought to adopt delicately.
I offer an alternative, but leave it to you to weigh up the differences and choose
accordingly. That is one overarching form of argument at work in the book, subtle
as it may be. But it is not the only one. The other meta-argument is an indirect
argument for the powers ontology and the powers themselves. When you adopt a
metaphysic, you thereby adopt the ontological core around which it is developed.
In effect, one endorses the ontology if one prefers the picture of the world it can
be used to paint. This form of argument—selling the ontology on the basis of the
metaphysic—is known as an ‘argument-by-display’.⁵ Hence, by situating the

⁴ I also believe it to be false. But that is beside the point, which is why mention of it has been stuck down
here.
⁵ Bigelow (1999: 45).
   

powers ontology within the powers metaphysic, I not only present a metaphysic
for your consideration, I provide an indirect defence of the powers ontology at
its core.⁶
I have three reasons for approaching the defence of the powers ontology in this
way. Firstly, I am not convinced that this is a debate best conducted in the trenches,
offering direct arguments for the existence of powers via pointed criticisms of neo-
Humean ontology. This is good work if you can get it, but without an alternative it is
unclear what those criticisms add up to. Moreover, a metaphysic is a grand thing, and
it is tough to walk away from one, even if you want to. It is, after all, the foundational
theory by which you structure your understanding of the world. And if I am right
that each metaphysic is mostly a package deal, wherein your foundational ontology
constrains your metaphysical options, then the majority of changes you can make
are going to be wholesale changes, which are not easily made. Hence, though there
is already a growing literature on powers that tackles the neo-Humean ontology
head-on, adding to it will not have the desired result unless coupled with a viable
alternative. And even then it may not succeed without recognizing the wholesale
nature of each.⁷
Furthermore, there is no denying that neo-Humeanism is a decent and coherent
metaphysic. Like your parents’ vacuum, it does a pretty good job. And so even if close
inspection identifies further flaws, they will be overlooked in light of its general
success. Hence, the advancement of metaphysics is better served by the development
of a serious alternative than it is by stockpiling bigger and bigger weapons to be used
against neo-Humeanism.
With that in mind, I do very little to respond to neo-Humeanism directly. Instead,
I present the key features of the neo-Humean metaphysic so we can know it when we
see it, appreciate its interconnections, and recognize how it differs from the powers
metaphysic. Along the way I point out some of the potentially unsavoury things to
which a neo-Humean might be committed, and how aspects of the metaphysic
might lead to certain kinds of problems (§2.6 especially), but I do not offer detailed
arguments to that effect. Dare I say, little is to be gained by rehearsing arguments
already in the literature: that is a waste of ink, even if some of the arguments are quite
good. And though there might still be ground that can be taken from the neo-
Humean by engaging with her views directly, the neo-Humean metaphysic is
unlikely to be deposed in this way.
My second reason for offering an indirect argument for the powers ontology is, as
George Molnar attests, that some of the strongest reasons for accepting an ontology
come from the metaphysical work it can do.⁸ Hence, rather than adding to the first
order, there is much to be said for developing a second-order argument for powers,

⁶ I thereby find myself in good company: Ellis (2001), Molnar (2003), Heil (2003, 2012), Bird (2007a),
Chakravartty (2007), Martin (2008), Mumford and Anjum (2011), and Vetter (2015) are among those who
devote space to developing a metaphysic with powers. See Jacobs (2017) for a collection of articles offered
in that vein.
⁷ Bernstein (2013) suggests that the powers metaphysic will not be taken seriously until it connects itself
with the dominant neo-Humean framework. I deny that this is possible.
⁸ Molnar (2003: 186).
  

and thereby engaging a second wave of challenges to the status quo. Hence, I hope to
show the value of the ontology by utilizing it when doing metaphysics.
Quite frankly, the powers ontology is so foreign to many contemporary thinkers
that exploring what sort of picture can be developed with it is key to assessing its
strength and viability. The powers theorist has a different starting place, a different
set of assumptions, and, most importantly, a different set of tools. From this new
ontology comes new problems and new answers to old problems. In fact, with new
tools even the old problems start to look new. We thus get a reworking of meta-
physics in light of this change in thinking—a metaphysics reboot, if you will.⁹ We see
this happening in some recent works within the powers metaphysic. Brian Ellis offers
a series of first-order attacks on neo-Humeanism, but rounds that out with a second-
order account of natural kinds.¹⁰ Mumford and Anjum use an ontology of causal
powers in developing a unique theory of causation; and Barbara Vetter adopts a
similar jumping-off point in championing a powers-based theory of possibility.¹¹
There are others, and there are sure to be more.
Of course, in order to get to the point of doing metaphysics with powers I need to
make clear what exactly they are. Not only because they are largely foreign, but
because even friends of powers disagree about their nature. This is relatively new
territory: neighbourly turf wars are to be expected! Hence, the reader will find that
I spend time presenting my take on the powers ontology, but the disputes in those
sections are friendly in-house debates, not attempts to show that the neo-Humean
metaphysic is broken.
This brings us to the third reason for the indirect approach: when it comes to
systematic metaphysics (those that attempt to solve a range of problems via repeated
applications of some core ontological entity or entities), indirect arguments of this
second-order variety are not only strong arguments for an ontology, they are the best
game in town. Our picture of the world as given by the scientific and manifest images
(including intuitions about how to answer various metaphysical questions) are tested
against the sorts of answers the fundamental ontology can be used to generate.
Through negotiation of these two vantages we arrive at a picture of the world that
gets the highest overall score. Thus, Molnar may have understated the importance of
a metaphysic in defending an ontology: systematic metaphysics and indirect argu-
ments go hand in hand.
Systematicity in metaphysics is to be contrasted with a piecemeal approach,
wherein metaphysical problems are solved one at a time, in isolation, without any
attempt to apply a solution proffered in one case to further problems.¹² The benefits
of approaching problems in a piecemeal way are twofold: first, one avoids the
philosophical handcuffing that comes with retrofitting an ontology for a new

⁹ But we should not overstep our bounds, as Bird (2016) argues has been done. It is one thing to apply
the ontology to a range of problems, but it is quite another to think that the ontology itself offers direct
solutions to these problems, or that no other metaphysic is capable of offering similar indirect solutions.
¹⁰ Ellis (2001). ¹¹ Mumford and Anjum (2011), Vetter (2015).
¹² I do not mean to suggest that piecemeal metaphysics is an intentional activity wherein systematicity is
actively avoided. Though some theorists might have principled reasons for avoiding systematicity, more
often than not piecemeal metaphysics is simply the failure to engage with metaphysical problems
systematically. See Hawthorne (2006: vii).
   

problem, rather than navigating the specific contours of the problem and
custom-building to fit; and second, a custom-built ontology may avoid unforeseen
errors or difficulties that plague the original. But these benefits—creativity aside—can
be more costly than they appear. The restrictions that working systematically
imposes are not without a basis. Rather, they are a corollary of treating the world
as having a single unified ontological structure, which is nothing more than any
sensible form of realism about the world should demand of its ontology. Systemati-
city also ensures that one’s ontological debtors are paid only once. To knowingly cast
aside systematicity in favour of a piecemeal approach runs foul of these realist
presuppositions. This is not to deny that there are certain advantages in doing so,
but they may prove to be short-lived and artificial. Hence, new and potentially
ingenious piecemeal solutions ought to be held accountable to systematic standards.
David Lewis famously adopted a systematic approach to metaphysics. His was the
incredible claim that ours is but one of innumerably many causally isolated concrete
universes: we are occupants of one universe within a vast pluriverse. How could one
hope to defend such an outlandish claim? The answer is that you step away from the
incredible claim—a first-order claim about what there is—and show its value in
second-order debates regarding what it can do. First-order debates are debates about
fundamental ontology. Second-order debates, on the other hand, concern the sorts of
metaphysical questions with which Lewis engaged, regarding the laws of nature,
possibility, causation, and the like. In providing satisfactory responses to these
questions he built a case for taking his fundamental ontology seriously.
Lewis denies having set out to do metaphysics systematically, and might well have
preferred to approach problems in a piecemeal way.¹³ There is, after all, something
attractive about pretending there is nothing more to the world than the problem
with which one is currently concerned. That artificial isolation permits solutions
that often do better by our intuitions than any systematic solution could; to wit,
systematic solutions can be entirely at odds with intuition. But this benefit of
isolated solutions is lost when they are applied elsewhere, on some other problem.
Lewis admits to having found himself repeatedly applying the same basic frame-
work to every problem he approached, starting with the same set-up and showing
how it could be developed to solve the problem at hand. Though bested here and
there by this or that piecemeal solution in isolation, when applied to a wide range of
problems the systematic solution scores higher on average and overall. This is the
manner in which a systematic solution earns it keep.
It is via this approach that Lewis defends the existence of the pluriverse, however
implausible its existence might initially appear. It is the same methodology David
Armstrong uses in arguing for a robust account of laws as universals.¹⁴ Taken at face
value, laws like this are unattractive postulates. But once put to work, one can see how
they can earn their keep, and this is the reason Armstrong endorses their candidacy
as members of the correct fundamental ontology. My aim is to enact this same

¹³ The same appears to have been true of David Armstrong. See Mumford (2007: 95).
¹⁴ Armstrong (1978, 1997, 2010). I understand Armstrong as offering a distinct metaphysic from both
neo-Humeanism and the powers metaphysic.
  

systematic-style project as it pertains to powers, simultaneously claiming that this is

the right way to think about the powers debate in contemporary metaphysics.¹⁵
But this is not the only reason why an advocate of systematic metaphysics would
employ an argument by display, nor is it the only reason I employ it here. The other
reason has to do with methods of reduction and reductive analyses that a systematic
metaphysic employs. A reductive analysis is one that provides an alternative set of
concepts by which some target concept can be understood. To be successful, an
analysis must be non-circular, and given in terms of concepts that are themselves
better understood. We thereby ‘tame’ the analysandum, replacing it, as it were, with
homier concepts. There is, metaphorically speaking, a downward direction here:
concepts with which we are less familiar are replaced by those with which we are
more comfortable.
As the concepts shift downwards, the entities with which those concepts are
connected start to move, too. That is, replacing one concept with others puts us in
an excellent position to do the same with the entities involved: the would-be reducer
of concepts is likewise the would-be reducer of the corresponding entities. Ideally, the
entities associated with the analysans can play whatever role we might otherwise have
required of the entities associated with the analysandum. One no longer needs to
countenance the reduced entities in one’s ontology; the reducing entities do the work.
Or, for those less inclined towards elimination, this manoeuvre renders the reduced
entities second class, expunging them from the fundamental ontology, but not from
the world altogether.
However, whereas the downwards direction of analysis leads to increased famil-
iarity (the analysandum is replaced by more familiar concepts), downward move-
ments in the ontological side do not have the same effect. Reduced entities are tamed
by the reduction—their reduction means that they are unpacked, demystified, elim-
inated, or the like—but the reducer is not similarly tamed. In fact, the ontological
reduction moves us towards entities that are thereby less well understood, because
they themselves are not similarly reduced.
The reduced entity starts out as a ‘black box’ whose mystery is solved via the
reduction. But the exposed mechanism within that black box introduces a new set of
cogs and pulleys, and thus a new set of mysteries or black boxes. We can solve those
mysteries, too, but only through further downward moves. But what if further
downward moves are not forthcoming? In that case the reducing entities constitute
the bottom of the ontological structure. They are a primitive of the ontology. Initially
this sounds like a bad thing. We have demystified one sort of entity only to replace it
with one or many entities which are themselves equally mysterious, if not more so.
How is this progress? Do not despair. Recall that the reducing entities—when they
are themselves incapable of reduction—are entities with which a host of problems are
to be solved. They are theoretical entities, postulated to do serious metaphysical
work. So, we have clearly gained in one sense. Further, we have reduced the number
of mysterious entities through the work they do, which is another plus. Nonetheless,

¹⁵ In contrast to those who would cast the debate as a local debate about properties. See Schaffer (2005)
and Wang (2016).
      ? 

it is clearly not desirable that one have genuinely mysterious entities at the heart of
one’s ontology. So, how does one make the fundamental entities less mysterious,
given that further reduction is out of the question?
The answer is that the aspects of the fundamental ontology get understood via the
very same procedure by which they are defended: systematicity. That is, one can gain
greater understanding of these fundamental entities by situating them in a wider
metaphysic and showing what they can do. We cannot open a cog, but we can see
where the cog stands in relation to other cogs, how it turns the pulley, and how it is
turned by another. And we can see what many such cogs and pulleys are jointly
capable of. This is how we demystify the fundamental and irreducible aspects of our
ontology.
Hence, my aim in situating the powers ontology within a wider metaphysical
framework is not only to argue that the powers ontology is worthy of adoption, but so
that more can be learned about powers and the powers ontology in the process. It is
only when an ontology is at home in a metaphysic that we can really start to see what
it is like. That said, total understanding is likely to remain beyond our reach, because
you can only fully understand something when you can get inside it and pull it apart.
This cannot be done for the fundamental entities, and so learning what they are like
by finding them a home is the best we can hope to do.¹⁶
There is, of course, no reason to think that there is just one metaphysic for each
ontology. (An ontology lists the raw ingredients of existence, but many dishes can be
prepared from those same ingredients by adjusting the recipe.) Nevertheless, there
might still be a best metaphysic for a given ontology (where what counts as best
depends on what virtues are in play). My aim is to develop part of a powers
metaphysic, but it is too early to say if it is part of the best. Getting to that point
requires taking the powers ontology seriously when doing metaphysics. Once that
step is taken we can develop rival metaphysics from the same ontological base. That
stage of metaphysical theorizing is the philosophical equivalent of Kuhn’s ‘normal
science’, wherein practitioners operate within a shared paradigm.¹⁷ For the time
being, the shared metaphysical paradigm is neo-Humean, and most debates are
internal to it. We are thus in a phase of ‘normal metaphysics’. Hence, before we
can hope to find the best metaphysic for the powers ontology, the paradigm needs
shifting. And doing that takes nothing short of a revolution.

1.3 The Start of a Revolution?

In what proved a major development in the debate over whether disposition ascrip-
tions are analysable in terms of stronger-than-material conditionals, C. B. Martin
argued that the proposed ‘simple conditional analysis’ fails due to the possibility of
finks.¹⁸ The simple condition analysis (SCA) tells us that:

¹⁶ This is why I find Barker’s (2013) criticism of powers unreasonable; he wants to see the insides, but
there are no insides to be seen.
¹⁷ Kuhn (1962). ¹⁸ Martin (1994).
  

[SCA] Something x is disposed at time t to give response r to stimulus s iff, if x were to undergo
stimulus s at time t, x would give response r.¹⁹

A fink is a device that senses the occurrence of the stimulus that would otherwise
trigger a given disposition, and quickly acts to add a disposition if it was absent, or
remove it if it was present. In the scenario Martin describes, a dead wire is made
finkish by attaching it to a device that adds current when the wire is touched. In those
cases where a regular (non-finkish) wire would conduct electricity, the device livens
the dead wire so it then conducts. In this finkish case, the wire lacks the disposition in
question (being live), but the associated conditional (if the wire is touched by a
conductor, then electric current flows from the wire to the conductor) is nevertheless
true, owing to the actions of the fink. (The dead wire will conduct if touched.)
Operating the fink on the reverse cycle renders the conditional false even when the
disposition is present. Hence, the analysis fails.
What Martin then proposes—perhaps prematurely—is that the reductive project
be abandoned altogether and that we embrace an ontology of irreducibly dispositional
properties (a powers ontology). Having considered Martin’s rejection of the SCA,
Lewis writes:
Once we scrap the simple conditional analysis, what should we say about dispositions?
Martin’s own response is radical: a theory of irreducible dispositionality . . . Since disposition-
ality is irreducible, it is not to be explained in terms of the causal and nomological roles of
properties, but rather vice versa.
Those who are disappointed with the usual menu of theories of lawhood and causation
might do well to try out this new approach. But those of us whose inclinations are more Fabian
than revolutionary, and who still back one or another of the usual approaches to lawhood and
causation, may well suspect that Martin has over-reacted. If what we want is not a new theory
of everything, but only a new analysis of dispositions that gets right what the simple condi-
tional analysis got wrong, the thing to try first is a not-quite-so-simple conditional analysis.²⁰
Unsurprisingly, Lewis follows his own advice, offering in place of the failed analysis a
revised and much more complex analysis. Unfortunately for him, though his revised
conditional analysis successfully evades the problem of finks, it is susceptible to a
different problem, that of antidotes (another class of manifestation preventers).²¹ Yet
further modified versions of the conditional analysis have been offered since, but we
can save ourselves time by skipping past the details and cutting to the chase: as yet no
wholly satisfying analysis of dispositionality has been located.²² There are a number
of reasons why this is the case, but the one that matters most to friends of powers is
that the proposed analyses have been reductive in nature.
Though ostensibly about mere disposition ascriptions, there is much more to the
Martin/Lewis debate than just the question of when we are warranted in applying
certain predicates to objects. That is to say, it is a short step from talk of disposition

¹⁹ Lewis (1997). ²⁰ Lewis (1997: 148).

²¹ Bird (1998). See also Kvanvig (1999). We will revisit the problem of antidotes—albeit for a somewhat
different reason—in Chapter 6.
²² Those interested in the cottage industry of attempts to analyse disposition ascriptions via counter-
factual conditionals are advised to start with the summary in Choi (2012).
      ? 

ascriptions to that of what sort of properties objects have. Hence, though the debate is
posed at the level of conceptual analysis, the real meat is at the ontological level. The
reason this debate can shift so quickly from predication to properties is simple
enough: the proposed analyses are reductive. A successful conceptual analysis
shows how one or many concepts can stand in for another without loss, where the
stand-ins are preferable, because they are better or more easily understood. And once
we have a successful replacement of concepts, it opens the door to an ontological
reduction. That is because we can then limit our ontological commitment to just
those entities corresponding to the stand-ins. In other words, the proposed analysis
would allow us to make sense of dispositionality using concepts that do not them-
selves rely on the notion of dispositionality to be understood. Once this substitution
is accepted, we are no longer required to take seriously the existence of sui generis
dispositions, just as long as we endorse the existence of the entities picked out by
the stand-in concepts. Or, as a weaker variant, we are then in a position to treat
dispositions as non-fundamental, as long as we endorse the fundamentality of the
reducing entities. Either way, dispositions suffer a diminished status: they are at best
properties non grata.
As I say, there is more at stake in this debate than mere predication. Martin’s shift
in the debate—from talk about conceptual analysis to that of properties—makes the
stakes clearer: there is ontological territory to be won or lost. This attitude is mirrored
in Lewis’s response: he wants to rescue the analysis precisely because he is concerned
with its ontological implications. But the stakes and ramifications do not end there;
there are metaphysic-level implications, too. Lewis recognizes that the downfall of the
conditional analysis might mean surrendering his accounts of laws, of causes, and of
many other aspects of the metaphysic he endorses. This explains why there is more at
stake than disposition ascriptions alone, but it does not yet explain why a powers
theorist would be sceptical that a satisfying analysis can be found. The answer, quite
simply, is that powers theorists deny that dispositionality can be removed from our
fundamental ontology. And this is where things really start to get interesting.
Lewis, recall, describes Martin’s suggestion that we abandon the reductive project
as ‘revolutionary’ and ‘radical’, and goes on to claim that adopting Martin’s sort of
approach to the nature of properties would require us to develop ‘a new theory of
everything’. Given the intimate ways in which metaphysical theories are connected
and the centrality that a treatment of properties has within a larger metaphysic (not
to mention what such an account of properties would mean for issues of laws, causes,
and modality), he is surely correct: following Martin’s proposal would require a
massive upheaval of the metaphysical status quo. Lewis warns us against such a
revolution, citing his Fabian preference to stick with what he later claims are ‘widely
shared ideas about properties, causation, lawhood and counterfactuals’.²³
It cannot be denied that Lewis’s ideas (or ideas near-enough-Lewisian) about the
aforementioned metaphysical matters are widely held: they constitute a key part of
the neo-Humean metaphysical framework that dominates contemporary thought.
But that alone does not make them correct, nor is it a good reason to subdue any

²³ Lewis (1997: 148).

  

uprising that may be in the offing. When it comes to questions regarding the nature
of properties, causation, laws, or all that is built upon the metaphysical foundation
they form, we should not adopt Lewis’s ideas without at first having considered
what alternatives are available. After all, deep dissatisfaction with the metaphysical
status quo may be a minority view, but there are known difficulties with the neo-
Humean framework that are often swept under the carpet. I thus suggest we enact
the metaphysical revolution that Lewis fears, if only to see what is left when the
dust settles. If the neo-Humean framework deserves to keep its place, then it
should have no problem withstanding such a challenge. But I advise against placing
any bets just yet.
The opening shot of that would-be revolution has already been fired. Martin has
proposed that we give up on trying to make sense of dispositionality in terms of
causes, laws, and inert qualities, and that we instead invert the metaphysical frame-
work so that dispositions (powers) are the ontological foundation. And more volun-
teers are starting to sign up. Like with so many revolutions, we seek to challenge the
neo-Humean orthodoxy with something new, but born of what was there long before
the neo-Humean orthodoxy took hold. Just as the neo-Humean orthodoxy owes
much to its Humean origins, any metaphysic that countenances irreducible powers
owes something to its Aristotelian ancestry, the very thing Hume fought to under-
mine. On the face of it, this might look like this is a case of out with the old and in
with the older, but just as neo-Humeanism is not merely Hume warmed over,
adopting a metaphysic based on an ontology of irreducible powers is much more
than Aristotle exhumed.

1.4 A New Theory of Everything

Lewis warns us that inverting the neo-Humean’s ontological structure will result in ‘a
new theory of everything’.²⁴ He is, for the most part, correct: things will not be the
same once turned on their heads. The core ontology will be radically different, as will
the totality of any metaphysic built around it. But it is too early to say whether
everything that makes up the powers metaphysic must also be different. However, as
finding out might require having the whole metaphysic in front of us, the answer will
have to wait for another time. Therefore, as I start to develop the powers metaphysic,
it is the differences that will rule the day. The differences du jour concern the nature
of causation and the nature of persistence over time.
The aim of this book is thus twofold: on the one hand the book is an attempt to
operate within a different metaphysic to that which presently dominates the philo-
sophical landscape. The second aim of the book is no more modest than the first: it is
to develop the parts of that new metaphysic that concern the nature of causation and
how fundamental objects persist. It would be very difficult to provide the entire
powers metaphysic within a single volume, but focusing on the matters of causation
and persistence gives us insight into much of that grander metaphysic, and just how

²⁴ Lewis (1997: 148).

      

different it is to the status quo. Regarding the theory of persistence, the account
promises to do more than is typically done in such accounts, by offering an
explanation of persistence. To that extent, the powers-based account of persistence
has something to offer friends and enemies of temporal parts alike. The account of
causation is similarly innovative, restoring genuine potency to the world. As we shall
see, drawing on the resources of the powers ontology makes available solutions
previously unavailable or inaccessible.
The powers metaphysic has a great deal to offer, but you might be hesitant to take
that first step. With that in mind, here is an initial reason for taking powers seriously
(it also allows me to set the record straight regarding the force of that reason). The
reason is this: physical theory describes the fundamental physical entities exclusively
in terms of what sound like powers. When physicists describe the fundamental
physical entities, they use such terms as ‘spin’, ‘charge’, ‘mass’, and so on. These
terms look to be power terms: they concern what the fundamental entities can and
will do. For instance, ‘charge’ names the power to produce electromagnetic fields,
‘spin’ the power to contribute to the total angular momentum of a system, and ‘mass’
the power to produce gravitational force.²⁵ Furthermore, not only do physicists use
power terms to characterize the fundamental physical entities, they only use power
terms: none of the ascriptions they apply look the least bit categorical.
Why does this matter to the present debate? The neo-Humean takes all properties
to be inert categorical properties. She might occasionally employ power terms, but
these terms name second-class properties. They are either a shorthand for the
categorical properties that are used in a proper analysis of them, or they are terms
we use until we know more about the genuine categorical properties at work. Thus,
the neo-Humean does not take her use of power terms to carry any serious commit-
ment to the existence of powers. But if it turns out that physics—the science
concerned with the most fundamental entities—continues to make use of power
terms, then the neo-Humean’s placeholder strategy runs into trouble. She might
maintain that these so-called ‘fundamental’ entities are not fundamental after all, and
that they, too, will give way to yet more fundamental categorical structures, but this
runs counter to what science suggests. Attempts to locate structure within the
fundamental entities have consistently come up empty, well beyond the point we
might have expected to find some.²⁶ It starts to look like some properties—maybe all
of the fundamental ones—are powers.²⁷
As it turns out, this last claim is too hasty. Despite the willingness of friends of
powers to declare the battle won in their favour, this fact about scientific practice does
not win them the day.²⁸ Physicists employ power terms—that is for certain. But two
points stand in the way of this settling the debate. First, these are the only descrip-
tions available to them. As Frank Jackson indicates,

²⁵ Ellis (2002: 47) and Mumford (2006). See Livanios (2017) for a dissenting opinion.
²⁶ Molnar (2003: 133).
²⁷ On a sparse conception of properties, all properties are fundamental. Hence, on that view, all
properties will be powers.
²⁸ I discuss this matter in detail in Williams (2011a).
  

when physics tells us about the properties they take to be fundamental, they tell us about what
these properties do. This is no accident. We know about what things are like essentially
through the way they impinge on us and on our measuring instruments.²⁹
And second, the terminological usage does not dictate the ontological facts. Even if—
as it appears—these power characterizations mark the end of scientific explanation in
microphysics (owing to the nature of its methodology), this alone does not fix the
ontological picture. It could still be the case that these properties are categorical, even
if we are unable to characterize them as such.
Try as you might (and many have), you cannot properly reason from facts about
terminological use in microphysics to facts about powers. Physical theory is, unfor-
tunately, more ontologically innocent than it appears. But even though this does not
win the day for the friend of powers, it ought to have the neo-Humean sweating
under the collar. She lacks a non-powers-based description of the fundamental
properties. The ontological question is an open one, but there are minor theoretical
advantages to treating the fundamental properties as being as described. If the neo-
Humean wants to deny that what you see is what you get (or more correctly, what
you infer from certain test behaviours is what you get), then she needs to offer
reasons in defence of her hidden categorical structure. Perhaps, in time, she will. Even
if she can, we nevertheless have a good reason for looking beyond the neo-Humean
status quo and taking the powers metaphysic seriously.

1.5 Ontology Fundamentals

There are many ways of developing a powers ontology, and many ways a friend of
powers might understand the natures of the powers themselves. Thus, if you are
relatively new to powers, you should keep in mind that mine is but one example: the
friends of powers is a club with an increasing number of members, and we do not all
see eye to eye on matters ontological or metaphysical. What you will find in this book
is one way that a powers ontology can be configured, along with the requisite
domestic squabbling and territory-marking involved in defending my preferred
configuration. That goes double for the metaphysic developed with it.
That said, the differences between the various configurations only look significant
from nearby. From the much removed neo-Humean perspective, they will appear
similar. Even as one gets closer, it is clear that the range of plausible powers
ontologies form a close-knit family developed with a similar vision and a shared
set of primitives. As with neo-Humeanism, one finds a similar flavour running
through them all. Thus, though the ontology on offer is (unsurprisingly!) what
I ultimately believe to be the correct one, and it has quirks that are peculiar to it, it
nevertheless ought to be sufficiently representative.
Regarding the details of the ontology, the powers are the star of our show, and as
we move forward they will hog the spotlight accordingly. But they are not the only
member of the cast: the supporting cast consists not only of the parts of the ontology
that get lower billing, but also certain meta-ontological assumptions on which the

²⁹ Jackson (1998: 23).

   

ontology and metaphysic rely. Some of those will be specific to the present view, too.
This last section provides a brief rundown of the supporting cast. Those interested in
these topics in their own right will find them rather terse, but spelling them out in the
full detail they deserve would take us too far afield.
Properties. Powers are properties. That is, they are ways that objects are. Objects can
be red, objects can have a mass of 0.0034 kilograms, and objects can be attracted to
other objects with like charge; these are all ways that objects can be. Moreover,
properties are repeatable ways that objects are, meaning that many objects can be the
same way. Many objects can each be exactly 0.0034 kilograms, or red, or attracted to
like-charged objects, and so on. These ways that objects can be constitute a funda-
mental ontological category. Properties are thus not reducible to sets or classes of
exactly resembling objects, nor to mere predications, or any other nominalist reduc-
tion. Put bluntly, I take properties to be real. This should come as no surprise: an
ontology that puts powers at its core ought to be one that treats those powers as
genuine existents.³⁰
I do not, however, take a stance on the matter of property similarity. Property
instances are genuine entities, and properties constitute a category of being distinct
from the object category—the category of property bearers—but I do not here
endorse the realist thesis that two instances of the same property type are strictly
numerically identical (universals) over the theory that property instances are par-
ticulars, pairs of which might be exactly similar (trope theory), or vice versa. Both
theories take properties to be real, and both afford an account of property and object
similarity, and that is sufficient for the ontology I develop.³¹
Furthermore, as ways objects are, properties are not substantial particulars in their
own right. Thus, the properties—those ways that an object might be—cannot, in and
of themselves, be the sorts of things out of which objects are composed. Thus, they
are not tiny building blocks of reality, or anything of the sort.³²
Objects. Objects are the things that bear properties: they are the things that are the
‘ways’. Objects may be thought of as particular substances (a particle; a field; an
electron), as some underlying ‘I know not what’ (substratum), or as particularized
natureless ‘property-havers’ (bare particulars), but I reject the claim that objects are
themselves nothing more than bundles or collections of properties (bundle theory).³³
There is an object category, and it is not reducible to mere collections of properties.
This is the flipside of the account of properties just presented. Properties are
ways. That phrase is not the clearest, but it gets across the main idea that objects
are capable of taking on a variety of forms. They come in this form or that, and can
change forms throughout their lifetimes. Those different forms are the ways that

³⁰ For a different tack, see Unger (2006). Unger defends a (non-trope) nominalist account of powers.
I am uncomfortable with this, to say the least.
³¹ See Tugby (2013a) for reasons to prefer universals, and Platonic realism in particular.
³² Pace Campbell (1981, 1990). Any trope theory would thus have to be more like Martin’s substance-based
trope theory (1980). This will prove significant to my powers-based accounts of causation and persistence.
³³ See Black (1952).
  

objects are. But you cannot have a way without a something that is that way. The
objects are the somethings.³⁴
In order to take seriously the suggestion that objects are nothing more than
bundles of properties, one would have to take properties to be sufficiently substantial.
They must be robust enough to be able to enter into building relations with one
another such that they can form objects. This makes properties more like mini-
substances: building blocks from which objects are composed. But I do not see
properties as at all like this.³⁵ So, what are objects, if not bundles of properties?
Beyond saying that they are something more—the something that has the properties;
the something that is the ways—I have little to add. But (I believe that) everything
I have to say about powers, and their role in the powers metaphysic, works equally
well under any of the standard non-reductive treatments of objects. I take the object
category to be real and fundamental. I likewise reject those ontologies that take
processes or ‘goings-on’ to be fundamental, such that objects are second-class
existents, abstracted away from those processes.³⁶ There are processes to be sure
(they are protracted events), but they depend on the objects and properties that
comprise them, and not the reverse.
Property instances are dependent entities because they depend for their existence
on the objects that instantiate them. But the dependence runs in the reverse direction,
too. There is no sense to be made of an object that is not some way(s). However
limited or expansive that range of ways might be, and however banal or unusual
those ways are, any object that exists must be some way(s). It must, for instance,
occupy some region, and differ from the region that it is in and from those around it,
and it must have capacities for how it will be soon after. It is impossible that any
object should fail to be any way at all, and thus all objects are such that they depend
on their properties as much as the properties depend on them. (My claim that objects
cannot exist ‘unclothed’ should not be read as a rejection of substratum or bare
particulars. I take both to be live options. What I deny is that either should exist at
any time without properties.)
States of affairs. There can be no property instances without objects that bear them.
And there can be no objects without ways those objects are. We can thus conclude
that entities from these two categories come as a package deal: nature presents us
with objects bearing properties. These package deals are known as ‘states of affairs’.³⁷
These are the basic existential units of the powers ontology I defend. Like Armstrong,
I take the world, at base, to be a world of states of affairs; unlike Armstrong, I take the
primary case to be objects bearing powers.³⁸

³⁴ For contrast, see Brian Ellis (2009) for an example of bundle theory plus powers.
³⁵ Consider: if properties were independent entities, then one might reasonably expect to find them
‘floating free’ of other properties or bundles. Pace Campbell (1981), this does not seem to be the case.
³⁶ The champion of process ontology is Whitehead (1931); his acolytes are few and far between. That
said, I detect shades of process ontology in some recent accounts of powers, including: Mumford and
Anjum (2011), Heil (2012), and Marmodoro (2017).
³⁷ Armstrong (1978, 1997). I follow Armstrong in understanding states of affairs as instances of
particulars instantiating properties and complexes thereof, not as abstract objects.
³⁸ Whether or not this is sufficient to fix all the facts about the world—to wit, whether a totality fact is
needed, too—is not something I have the space to explore here. See Armstrong (2004a).
   

‘State of affairs’ is not a complicated notion. In fact, as presented, it resembles

objecthood, especially once we appreciate that objects cannot exist without proper-
ties. So, why not limit the ontology to objects? The answer is that not all states of
affairs are objects. Included as states of affairs are objects standing in relations to one
another. Armstrong himself understands a state of affairs to be ‘a particular’s having
a certain property, or two or more particulars standing in a certain relation’.³⁹ In fact,
relations play a key role in Armstrong’s argument for the existence of states of affairs.
Consider the asymmetrical relation loves, the fact that Joanie loves Chachi, and the
(lamentable) fact that Chachi does not love Joanie. The mere existence of Joanie and
Chachi and their instantiating the loves relation does not alone determine the right
set of facts in this case. With just those three constituents, three different sorts of facts
could obtain: (i) Joanie’s unrequited love of Chachi; (ii) Chachi’s unrequited love of
Joanie; or (iii) Joanie and Chachi’s mutual love of one another. We are after (i). That
means the world must be appropriately arranged such that Joanie stands in the loving
relation to Chachi, but not the reverse. The constituents alone are not up to the task,
but states of affairs can come to our rescue. Joanie’s loving Chachi (and no similar
loving relation going the other way) is a specific state of affairs. Hence the need for
states of affairs.⁴⁰
As the above implies, including states of affairs in one’s ontology constitutes an
‘addition of being’ to the world.⁴¹ This nebulous phrase means that something extra
must be included in the world in order to fix the facts. But that something need not be
some thing, if by ‘thing’ one means an object or a property. The extra something is
perhaps a structural feature, in the sense of ‘structure’ Sider employs, but it need not
be a concrete (or even abstract) entity of any sort.⁴² This addition of being stands in
contrast with Armstrong’s notion of the ‘free lunch’. To the layman, a ‘free lunch’ is a
lunch you do not have to pay for. For the ontologist, a ‘free lunch’ is a fact/truth/
entity that obtains/holds/exists in virtue of some more fundamental fact/truth/entity
obtaining/holding/existing (it can be any or all of these). For example, it is commonly
held that if the parts of an entity are properly configured, then the whole that those
parts form is something one gets for free.⁴³ Armstrong takes determinable properties
to be on the menu for the free lunch: if some object is a determinate shade of red,
crimson say, then it is true that something is red in virtue of that object’s being
crimson. The redness comes for free when crimson is instantiated; the redness is no
addition of being.

³⁹ Armstrong (1978: 80). Armstrong uses ‘particular’ where I would use ‘object’.
⁴⁰ Armstrong (1989: 17). As Armstrong notes, a trope theorist can avoid the need for these states of
affairs if her tropes are non-transferable, meaning that Joanie’s love of Chachi is essentially Joanie’s love of
Chachi, and could not have been the reverse, or a love of anyone else by or for anyone else. This fixes all the
relations without the need for states of affairs. For convenience, I will continue to speak of states of affairs,
with the recognition that a trope substitute is available for some of these.
⁴¹ Armstrong (1997). ⁴² Sider (2012).
⁴³ It is customary to think of free lunches as things of a higher level that depend on things of a lower
level. Armstrong analyses ‘free’ in terms of supervenience, but nothing prevents us from employing other
building relations. I myself am rather sceptical of free lunches, ontological or otherwise. Amusingly,
Armstrong once went to some lengths to buy me lunch, but I do not believe it was in an effort to prove
a point!
  

Sparsism. To be a sparsist—as opposed to an abundantist—about properties, is (at

least) to believe in the existence of only those property types required ‘to characterise
things completely and without redundancy’, where the characterizations are (typic-
ally) those coming from our best sciences.⁴⁴ Some versions of sparsism add further
restrictions; Lewis’s sparsism, for instance, adds that the sparse properties are those
that: underlie qualitative similarity, carve at the joints, are intrinsic, and are highly
specific, to name a few. Some of these are harmless restrictions, but let us put them to
one side and stick with just the minimal understanding.⁴⁵ On that minimal under-
standing, a sparse property type is one that a realist account of the natural sciences
cannot do without when characterizing the relevant objects. I am this sort of sparsist,
and will assume sparsism in what follows. (Why just the minimal interpretation?
I ultimately endorse maximal sparsism, but some of the additional restrictions step
on the toes of what I defend elsewhere in the book, and so are not restrictions I want
to assume without argument.)
The world needs to have certain properties in it to do all the things it does, and also
to account for what things it could or might do, and so our best account of the world
needs to make reference to those properties. But those same theories do not require
that the world have in it any additional properties beyond those just mentioned, and
so I reject abundantism. This minimal sparsism, then, is really just a kind of property
minimalism, in the spirit of Occam’s razor. It is a refusal to posit properties that our
best theories of the world do not demand we include.
Regarding those ‘best’ theories, I have presented them as coming from our best
sciences, and for the most part I take that to be true. In fact, I tend to further restrict
the set of theories to those just coming from our fundamental sciences (more on this
below), but we need to leave room in case there is more than this required to make
the world tick. Hence, I reserve space for something more. The more might well come
from decidedly non-empirical investigations into the nature of the world. That is, it
may come from philosophy, and from metaphysics in particular. Consider: it might
be that the account of the world as given by the natural sciences—or physics in
particular—is complete. But it just as well might not be. And not just because the
picture of the world as given by physics might miss out on essential aspects of
macroscopic phenomena (which it perfectly well might), but because even the fullest
final account that incorporates every scientific domain might still be incomplete. The
methodology of the sciences is such that some properties might simply be outside of its
scope. Take ethical properties, for instance. Assuming they are real and evade reduc-
tion, then there may be properties present in the world to which science has little or no
access. There may be similarly evasive metaphysical properties. And so I would rather
leave open the epistemic possibility that there are properties (however mysterious)
beyond those that science discovers, than legislate against them from the word go.
Micro-reductionism. I have said something about the general form that objects and
properties take, and I have indicated that I presuppose the sort of minimalism that

⁴⁴ Lewis (1986a). See Armstrong (1992) and Schaffer (2004).

⁴⁵ As Schaffer (2004) shows, there is room for disagreement regarding one’s sparsism.
   

sparsism entails. But I have not said much about which sparse properties there are,
nor anything specific about the objects that have them. The short story is that I pair
my property sparsism with a general property and object fundamentalism. That is,
I will assume (to some extent) that the properties and objects that exist are those that
we need to ground the truths about all other would-be objects and properties. In
other words, I pair my minimalism with yet more minimalism.
The basic assumption behind both is that we should make do with as little
ontology as we can. Naturally, this needs to be balanced with any rise in complexity
that comes with a reduction in ontology.⁴⁶ My fundamentalism is therefore defeas-
ible, and would be defeated if the theoretical cost of a minimalist ontology is too
great. But on the assumption that the theoretical costs are less than, or at least no
more costly than, any richer ontology, then we should refrain from postulating the
existence of entities that we can do without. Of course, ‘doing without’ is one very
large promissory note that will surely bounce if you try to cash it any time soon, but
the lack of present concrete answers and explanations regarding the ‘hows’ should
not dissuade us from wanting to explain less fundamental things in terms of those
which are more fundamental.⁴⁷
As previously indicated, what is fundamental might not simply come from the
picture of the world as given by our best future physics. It might turn out that people
are fundamental existents, or that the fundamental base is not an array of tiny
particles or fields, but a giant universe-sized field, or all of spacetime. That is, it is
not a requirement of fundamentalism that the fundamental entities be microphysical.
If there is just one (giant) fundamental entity on which all else relies, that would still
be in keeping with my fundamentalism, just as long as the other candidate objects
and properties are ultimately explained in terms of it.
That said, my bet is on the rather unoriginal thesis that the fundamental entities
will be those postulated in our best physical theses, along with whatever not-the--
subject-of-physics properties we need to postulate in order for the world to run.
Consequently, you will find me leaning towards a micro-reductive picture of the
world, when and if I can. And when I cannot—which is much of the time, owing to
the aforementioned promissory note—I happily make use of mid-sized ontology (it is
a better place to find homey examples). In the long run, that midsized ontology will
either give way to something smaller, or it will not. I am happy with either scenario,
as long as there is no ontological waste involved. I would be similarly happy to learn
that there is just one giant object, or that the only property bearers are fields. I will
continue, for the time being, to talk of particles and tables, with the recognition that
the latter will probably give way to the former, and that even the former might not be
the last stop in the descent to the universe’s basement.
For those who are worried about micro-reductionism, or worried about any form
of fundamentalism, I remind you that these theses alone do not entail the non-
existence of macroscopic phenomena. That is, it does not follow from the fact that

⁴⁶ Compare with Sider (2012).

⁴⁷ I leave the details up to the grounding theorists, or whomever replaces them in that particular
philosophical niche. See Correia and Schnieder (2012).
  

some entities are fundamental that the non-fundamental entities are non-existent.
One might combine micro-reductionism with a building relation that gets you
macroscopic entities for free.⁴⁸ Nor is it any part of the micro-reductionism that
I favour that all (or even the best) explanations of all phenomena will be couched in
those terms. Micro-reductionism is an ontological thesis concerning what entities
exist fundamentally; this implies, but does not entail, micro-explanationism.
Eternalism? I am not going to assume that eternalism is true. Eternalism is the view
that the temporal dimension is analogous to the spatial extension, wherein all times
(and the objects that exist at those times) are on equal ontological footing, and the
present is not ontologically privileged.⁴⁹ Just as you do not consider your present
location to be ontologically privileged—meaning that being spatially removed from
your present location does not affect the existence of those things so located (when you
are in John o’Groats, the cliffs at Land’s End are no more or less real than the couch on
which you are sat)—eternalists do not consider objects located at times earlier than or
later than the present to be any more or less real. Eternalism is standardly contrasted
with presentism, the account of time according to which the only things that exist are
those which presently exist, but can be contrasted with any account that privileges the
present, such as the growing block or moving spotlight accounts.⁵⁰
When push comes to shove, I consider myself an eternalist, but I will not here
assume that eternalism is true. Nor will I defend eternalism. In fact, eternalism does not
play a significant role in my ontology as far as the present discussion is concerned.
Those who have a problem with eternalism are thus invited to forget I brought it
up. So, why did I bring it up, you ask?
The first reason is that Chapter 8 offers a powers-based account of persistence, and
though the account is friendly to perdurantists, endurantists, and stage theorists alike
(and their preferred accounts of the metaphysics of time), the basic framework
favours the eternalist. (It is friendly to all, but a little more friendly to some.) The
second reason is going to sound rather wiggy. It is so that you can be on the lookout
for eternalist modelling. An ontology centred on causal powers is compatible with all
standard theories of time, yet one finds that many friends of powers favour the
combination of presentism and endurantism.⁵¹ I favour neither. But the point is not
that we have different beliefs about time and persistence, nor that these matters are
largely orthogonal to the present discussion, it is that it is useful to model powers in
broadly eternalist ways, and this is not the norm. I am thus putting my cards on the
table. Finally, should it turn out that the metaphysics of time and persistence are not
orthogonal to the best way of configuring a powers ontology, it is good to have had
the eternalism out in the open.

⁴⁸ Bennett (2011) makes use of a general class of metaphysical relations that she calls ‘building’ relations,
that includes grounding, dependence relations, and the like.
⁴⁹ See Smart (1963), Mellor (1998), and Sider (2001).
⁵⁰ See Cameron (2015) and Skow (2015). If one drops the requirement that all times be on equal
ontological footing (or conversely if one denies that the spotlight of the moving spotlight carries with it
ontological privilege), then eternalism and the moving spotlight could be combined.
⁵¹ See Mumford (2009a), Wahlberg (2009), and Heil (2012).
 2
The Neo-Humean Status Quo

2.1 The Humean Stain

The vast majority of contemporary metaphysics is pursued from within what is best
described as a neo-Humean framework. This is not to suggest that the ranks of
metaphysicians run deep with Hume scholars, nor that doctrines lifted directly from
the Treatise or the Enquiries hold sway in contemporary debate. But many of the core
metaphysical principles and ideals at work in the contemporary arena have a
distinctly Humean flavour, and can be seen as natural extensions of those that
Hume endorsed or assumed. I hazard that most of what passes for good metaphysics
these days bears some Humean stain, intended or otherwise. The same could easily
be said of the metaphysical aspects of contemporary philosophy in general. Hume’s
influence is ubiquitous.
Nowhere is this influence more evident than in the work of arch-neo-Humean
David Lewis. Lewis was the modern-day Humean par excellence, and unashamedly
so. The influence of Lewis’s work on contemporary metaphysics has been profound,
and for quite some time the debates at the forefront of metaphysics have been those
set by Lewis. Many of the popular solutions are his, too. Even the dissenters tend to
be more congenial to his solutions than not. I would not hesitate to declare Lewis the
most influential metaphysician of the last quarter of the twentieth century, and the
incumbent for the first quarter of the twenty-first; few would disagree. A by-product
of Lewis’s influence on contemporary metaphysics—if not constitutive of it—is the
widespread adoption of views consonant with (if not identical to) those he held. This
has only added to the prominence of the neo-Humean metaphysic.
The pervasiveness of this influence can be seen in the extent to which adopters can
be unaware of just how much their views borrow from the Lewisian-cum-Humean
corpus. Consider: given the ways in which metaphysical theses are interlaced,
conscious adoption of a Lewisian (or near-enough Lewisian) thesis brings with it
unconscious adoptions (not unlike the way little extra programs might be added to
one’s computer when loading a targeted program). Friendly dissension from those
same theses can have much the same effect, as the space of purportedly viable
alternatives is itself highly circumscribed: when you set the debate you thereby
determine the range of ‘acceptable’ solutions. Consequently, many debates in con-
temporary metaphysics are more like internal debates than they first appear, situated
within the Lewisian-cum-Humean framework and criticizing one or another aspect
of it, even when they are conducted in earnest, with an effort to keep assumptions
at a minimum. Indeed, even those apples that fall furthest from the Lewisian tree
still tend to fall within the greater neo-Humean catchment area. Solutions and

The Powers Metaphysic. Neil E. Williams, Oxford University Press (2019). © Neil E. Williams.
DOI: 10.1093/oso/9780198833574.001.0001
  -  

suggestions that deviate too greatly from that corpus are given short shrift for exactly
this reason, despite any independent plausibility they might boast. Any philosopher
worth her salt ought to be willing to adopt the conclusion of a deductively valid
argument based on independently plausible premises. But many of the so-called
‘independently plausible’ premises that pass muster in contemporary metaphysics
are developed within a neo-Humean framework, and hence we might question the
independence of those that appear to support neo-Humean conclusions.
Plenty of folks deny that they subscribe to anything neo-Humean, not even in a
distantly removed form. But the nature of metaphysics makes it hard to avoid
adopting various theses even when they are not one’s focal topic. Contemporary
metaphysics is ‘theory-laden’ in the Duhem–Quine sense: theories are connected to
one another, rely on one another, and are applied in groups.¹ Hence, you cannot pick
out this piece or that without bringing along another. Moreover, the most successful
metaphysics is systematic: its parts are interconnected in such a way that a few central
theses have wide application across the scope of metaphysics. And even when
systematicity is not the aim, it is not feasible to engage in metaphysical thought in
a way that legitimately isolates one metaphysical puzzle from all others. At any given
time a specific puzzle will take centre stage, but rough solutions to others will be
working behind the scenes, pulling the ropes of constraint and preference. Meta-
physics is all but impossible to do without either having a rough framework in mind
or developing one en route, even if that framework remains largely in the back-
ground. This is where the neo-Humean metaphysic hides when it is not out front for
all to see. I am thus claiming that most folks employ a metaphysical framework that is
largely neo-Humean—be it in the foreground or the background—even if its pos-
sessors are unaware that they employ one at all.
When we do metaphysics, we work with a background framework, however rough
and ready, and try to resolve this or that metaphysical puzzle from within it. In the
introduction to his first volume of collected papers, Lewis explains that it was not
his aim to be a systematic philosopher. He ended up one regardless, but notes that
he would have liked to have approached problems piecemeal, offering ‘independent
proposals on a variety of topics’. It was not to be. As he says, he ‘succumbed too
often to the temptation to presuppose [his] views on one topic when writing on
another’.² Metaphysics, quite frankly, is very hard to do. With so many moving
parts, progress is made easier by fixing as much of the framework at a time as one
can. Thus, I do not think Lewis can be criticized for having gone about things in
the manner he did. I doubt he could have made the progress he did had he done
otherwise. Perhaps the ways we reason make us incapable of doing anything else.
But this procedure is not limited to metaphysics. We have a background in place—
a metaphysical background—when we engage in most philosophical investigations.
This should not come as a surprise. One can hardly ask, for instance, whether
deteriorating artworks should be restored to their former glories, without thinking of
them as objects with mutable properties. As we start to find solutions, the background
becomes increasingly fixed. And even though ideas about the nature of properties will

¹ Duhem (1954), Quine (1951). ² Lewis (1986b: ix).

     

not be at the forefront of one’s thoughts on those occasions, how one envisions
them will place subtle restrictions on the sorts of solutions one can entertain. As will
one’s beliefs about change, time, possibility, objects, and so on. The background
metaphysic need not be well formed or ever at the forefront for it to exist or make a
difference: even a background impression of properties as inert qualities impacts
how one sees the place of artworks in the world. I will spare us all the tedious task of
showing the influence one’s background metaphysic can have on our thinking,
philosophical or otherwise; hopefully the point is clear enough.³ But even if it is not,
one might begin to appreciate the role neo-Humeanism plays in one’s own thinking
as I present the core of the account.
So, what exactly is this neo-Humean metaphysic that has contemporary meta-
physics in its grasp? I see it as having four main components, concerning: properties,
possibility, laws of nature, and causes. More specially, they are that: (i) properties are
inherently non-modal and passive; (ii) what is possible is restricted only by imagin-
ation and coherence; (iii) the laws into which objects enter do not force them or push
them around—they are non-governing descriptions of the positions and states the
objects take up; and (iv) causation is a similarly weak and extrinsic relation. The
components of the neo-Humean metaphysic overlap in various ways, but at their
heart is the doctrine of Humean Supervenience, according to which the most
fundamental or sparse properties are categorical, thus lacking modal or causal
characteristics, and so whatever causal or modal facts obtain do so because they are
determined by the ‘vast mosaic of local matters of particular fact’ (that is, the limited
base of categorical properties).⁴
What follows is an introduction to the neo-Humean metaphysic. I warn the reader
that this is not supposed to be anyone’s view in particular; there is no exact canon to
which all neo-Humeans subscribe. What typifies neo-Humean metaphysics is a
family of connected theses that play off one another in various ways. My aim is to
provide the metaphysical ‘flavour’ these views tend to share, thereby presenting some
of the neo-Humean theses that dictate contemporary metaphysical thought and that
lurk in the metaphysical dimensions of the rest of philosophy. The components of the
view I set out most clearly resemble those of David Lewis, but the point is no more
Lewisian scholarship than it is Hume scholarship.

2.2 Of Possibility and Necessity

A natural ancestor of Hume’s account of what is possible (and of what serves as an
adequate guide to discovering what is possible) is the first major aspect of the neo-
Humean metaphysic we will consider. According to Hume, if a proposition can be
properly entertained (a mental exercise of some sort), and it does not run foul of the
constraints of logical consistency, then it describes a genuine possibility. Hume thus
claims that conceivability is a guide to possibility. He famously states that:

³ I will, however, direct you towards Groff (2012), in which she provides evidence of the neo-Humean
stain in contemporary social and political philosophy and discusses the restrictions it imposes on the range
of viable views.
⁴ Lewis (1986b: ix). See also Carroll (1994).
  -  

’Tis an establish’d maxim in metaphysics, That whatever the mind clearly conceives includes the
idea of possible existence, or in other words, that nothing we imagine is absolutely impossible.
We can form the idea of a golden mountain, and from thence conclude that such a mountain
may actually exist. We can form no idea of a mountain without a valley, and therefore regard it
as impossible.⁵
For the most part, what Hume treats as possible is constrained by logic and
imagination alone. That means that the space of possibilities is particularly generous.
If you can dream of talking donkeys without contradiction (you can), then such
fantastical things are possible, despite their being non-actual. Round squares, by
contrast, do not fare so well, failing on both counts. Something’s being both round
and square is not only contradictory, it is unimaginable, too.
Just as Hume takes only ‘clear’ conceiving to be a guide to what is possible,
contemporary versions of this conceivability-driven approach to possibility place
various demands on the sort of conceiving adequate for picking out what is possible.
Some follow Hume, insisting that only deep or clear conceiving can tell us what is
possible; others place no limits on the operations of the imagination, but restrict the
role of conceiving to that of providing evidence for what is possible. Nonetheless,
even with those conditions in place, the range of possibilities remains vast.⁶ Talking
donkeys? Sure. Fire-breathing dragons? Why not. One-eyed, two-toed, three-armed,
vengeful giants? But of course, this is a catholic approach to possibility! There is
virtually no limit to the ways in which the constituents of the world could legitimately
have been combined, or how they can be recombined, or how all these can or could
have been arranged.⁷ One can limit the scope of possibilities under consideration for
some particular purpose or other (like if we were to consider only those possibilities
in which the laws of nature are exactly like the actual ones), but this does not restrict
the scope of what is possible in general.
A sister thesis, often considered part of that mentioned above, takes epistemic
possibility as a guide to metaphysical possibility. For a proposition to be epistemically
possible is for it to be such that (for some subject) it might be the case that the
proposition is true, for all that subject knows.⁸ A proposition is epistemically possible
for some agent just in case the proposition’s being true is compatible with everything
the agent knows (or seems so to the agent), even if not suggested by it. Hence, it is
often claimed that a proposition’s being such that it could be true for all we know is
taken to be sufficient for (or evidence for) its describing a genuine metaphysical
possibility. If, for all we know, gold could have been a thinking substance, then ours
could have been (or is) a world in which gold thinks. Thinking gold is thus, according
to this account, a genuine possibility.
We can now ask what sorts of things the neo-Humean takes to be possible, based
on what we can imagine. Pausing first to put on our Humean hats, let us return to the
talking donkeys. The ability to speak is, by and large, restricted to humans. But
imagine a human who cannot speak (this, unfortunately, takes very little imagin-
ation). We can imagine—without contradiction—a human lacking this ability, and

⁵ Hume (1978: I.ii.2.2). ⁶ See Yablo (1993) and Chalmers (2002). ⁷ Lewis (1986a: 87).
⁸ Chalmers (2011).
     

therefore what we have imagined is a genuine possibility. The two (being a human
and the ability to speak) are now metaphysically cleaved, because we can clearly
imagine cases of one without the other. Two things result from this mental cleaving.
First, we take our ability to cleave these two as evidence for (if not constitutive of ) the
fact that the ability to speak is not necessary to humans.⁹ Second, it adds being human
and the ability to speak to our cache of raw constituents that can be combined and
recombined to form new imaginings. Drawing on that recently expanded cache of
constituents, we can now combine the ability to speak with that of something’s being
a donkey. No apparent contradiction arises, and so again we have described a
genuine possibility: that of talking donkeys. ‘To form monsters, and join incongruous
shapes and appearances’, writes Hume, ‘costs the imagination no more trouble than
to conceive the most natural and familiar objects.’¹⁰
It is surely the case that no donkey has ever spoken, and it is unlikely that any ever
will, but according to Hume and the neo-Humeans it is nevertheless possible. Perhaps
this would have required evolution to have taken a different path, or Doctor
Moreau’s work to have received more funding, but it is deemed possible nonetheless.
Through two simple acts of imagining, one has undermined the claim that the ability
to speak is a necessary feature of humans, and recombined the constituents to
factually describe something that is not the case but could have been. The world is,
to borrow Hume’s phrase, ‘loose and separate’ in this cut-and-paste sense. Bits of the
world can be cleaved from one another in the imagination, and the imagination is
then free to attach them elsewhere in a variety of ways. There is virtually no limit to
the mental ‘collages’ you can construct from dicing your experiences into tiny bits
and sticking them back together in different ways. A few collages are off limits, but
most are fine. The range of logically consistent imaginable recombinations is vast,
and therefore so, too, is the neo-Humean’s space of what is possible.¹¹
Hume’s cases build upon the set of impressions gained from experiencing the
world. The ideas we have ‘are copy’d from our impressions, and there are not any
two impressions which are perfectly inseparable’. Consequently, any time ‘the
imagination perceives a difference among ideas, it can easily produce a separ-
ation’.¹² But the neo-Humean does not ‘limit’ herself to impressions alone. She
extends the range of what is possible to things we have never seen and that cannot
be formed through recombination of the parts of those we have. The neo-Humean
thus adds to the pool of available constituents a set of ‘alien’ properties: properties
not possessed by any actual entities (past, present, or future), nor obtainable by
means of conjunction, interpolation, or extrapolation, of actual properties.¹³ The
result is a staggeringly liberal account of possibility: you can construct a lot of
collages! Adopting this general approach to what is possible leads to a great deal of

⁹ It may be that our ability to imagine the separation of the two is a reflection of our believing that
having the ability to speak is not necessary to humans, but the order here is not significant.
¹⁰ Hume (1975: I.ii.13).
¹¹ See Lewis (1986a: 87–90). Lewis denies that imaginability is much of a guide to possibility, unless used
in the way I have described.
¹² Hume (1978: I.i.3.4.).
¹³ See Lewis (1986a: 159–65), Armstrong (1989), Heller (1998), and Divers (1999).
  -  

monstrous things which, according to the neo-Humean, could have been the case,
despite their not being so.
But the largely unconstrained cutting and pasting does not end there. So far I have
considered examples concerning the recombination of objects and their properties.
But the neo-Humean’s liberal space of possibilities extends well beyond just these
ontological categories. If Hume is known for anything, it is that there is nothing deep
to be discovered about the sequence of events, vis-à-vis their causal ordering, beyond
certain psychological projections we contribute. As much as one might take a pair of
events to be causally related such that one must precede the other, Hume believes
that if one can imagine the former having occurred without the latter, then such an
occurrence could have taken place. And though past experience might tell us that the
two events are constantly paired, that is all we have to go on when we consider
what—if anything—must follow from the former event. Once you have satisfied
yourself that objects, properties, and events, can be cleaved from one another in the
imagination, one has a hard time seeing why any entity should be so fixed that it, too,
could not be separated from the circumstances in which it was found, and, at least in
theory, placed anywhere else. Everything—that is, every thing—begins to look loose
and separate.
This sort of thinking about what is possible goes hand in hand with what is
referred to as ‘Hume’s Dictum’: the principle that there are no metaphysically
necessary connections between distinct, intrinsically typed entities.¹⁴ What is this
principle supposed to mean? To be intrinsically typed just means that what deter-
mines the sort of thing the entity is depends on nothing outside of it; that leaves us
with the rejection of metaphysically necessary connections between distinct entities.
A metaphysically necessary connection is a connection that could not fail to hold,
come what may. If one entity is necessarily connected to another, then nature dictates
that where you find the first (or when the first obtains, or is, or what have you), the
other will be there, too. This could be: the having of some property by an object; two
objects so connected; two events such that one event is always followed by the other;
and so on. But in every case the entities are ‘fixed’ together, so to speak. Hence, their
being separate in nature (as opposed to one’s imagination) would not be possible.
You might now wonder what it means for them to be distinct, given their fixity.
After all, if they are necessarily connected, in what sense are they distinct? The
picture gets clearer if we start with non-distinctness. Clearly non-distinct (that is,
identical) entities bear necessary connections to ‘each other’. (They are one and not
two, and so talk of connections between ‘them’ is artificial, but it ought to be easy to
grasp how facts or features of the ‘one’ would thereby necessitate the same in the
‘other’.) Now consider non-identical entities. Something about their being non-
identical opens up the conceptual space for difference. We might even say we can
imagine one without the other, and in so doing we see both the sense in which we
understand them to be distinct, and the source of the controversy. To believe in
necessary connections in nature is to believe that things we can imagine as apart

¹⁴ This formulation is due to Wilson (2010). Wilson (2015) presents the role of Hume’s Dictum in
Lewis’s system. See also: Hume (1975: I.iii.6), Lewis (1986a: 87–8), and Armstrong (1997, 2010).
     

cannot, in fact, obtain as such. That does not make them one, but rather means that it
is up for grabs whether the ways we individuate objects entail their being separable in
nature. The neo-Humean claims that this is so entailed: to be distinct is to be free of
necessary connections.
Hume applied Hume’s Dictum to the case of causal connections, but neo-
Humeans expand its application to virtually any putative connection between entities
of any ontological category. It is a neo-Humean rule of thumb that necessary
connections are prohibited. The ramifications this has on how the neo-Humean
understands laws and causes are straightforward: causal relata are loose and
separate—there is no oomph, in the sense that the cause and laws do not make
their effects come about necessarily—and could have been otherwise. What laws
obtain need not have done so. Similarly straightforward are the implications the
principle has for the powers-based metaphysic I develop: according to the neo-
Humean, there cannot be properties whose natures have it that their potential
manifestations are part of their essences, as any property (or law, or cause, or effect)
can be pulled apart from any other. Powers are properties that have possible futures
built into them, and hence they do not play well with Hume’s Dictum or the neo-
Humean account of possibility more generally. I will discuss the neo-Humean
approach to laws and causes shortly, once I have talked about the neo-Humean’s
thinking regarding the nature of properties. But before getting there I want to say
something more about the lack of necessity in the neo-Humean framework.
According to the neo-Humean, there is a great deal that is merely contingent in the
world. As indicated, if you eschew necessary connections, then you are going to
endorse accounts of laws and causes that permit the possibility of them being quite
different than they actually are. Likewise, your properties will be capable of attaching
themselves to those laws in all sorts of different ways. Things could have been really
quite different from how they are. Contingency abounds. But an interesting exten-
sion of this thesis gets applied—somewhat surprisingly—to metaphysics itself, as
I shall now explain.
One of the many ways in which the neo-Humean and I part company over matters
of possibility concerns the possibility of alternative metaphysics. Consider what we
might call the ‘basic metaphysical facts’ about the world. These facts constitute the
solutions to a great many metaphysical puzzles. They are, so to speak, rules about
how things must be, metaphysically. Take the debate over the structure of concrete
particulars. Some folks deny they have structure, others take them to be complexes
formed by uniting an underlying substratum with property instances, others yet
claim that they are nothing but collections of appropriately related property
instances. This last view is the bundle theory. Let us assume, for argument’s sake,
that the bundle theorists are right about the composition of concrete particulars.¹⁵
That is, we shall assume that bundle theory is true. This fact—the truth of bundle
theory—is what I am calling a ‘basic metaphysical fact’. Why this is a fact, and a
metaphysical one at that, ought to be obvious. But its being basic is less obvious. What
I mean by describing it as ‘basic’ is that it gives a rule regarding the construction of

¹⁵ They are not. But that is a discussion best had elsewhere.

  -  

concrete particulars, but does not describe any specific instance. A basic metaphysical
fact then is just a rule regarding the ways in which the world must be constructed.¹⁶
Any concrete particular that exists—if any do—must be such that it is a collection of
property instances, in accordance with the rule.
Let us assume that there is no controversy in claiming that there is a set of facts like
this in operation in the world. This is the stuff of long-standing first-order meta-
physical debate, and so ought to be as good as true. Nevertheless, there seems to be
some difference of opinion regarding the status of these facts. I claim that the basic
metaphysical facts are necessary; that is, any ways in which the world might have
been (or could have been constructed, if you prefer) must conform with those facts.
Neo-Humeans typically deny this, treating these facts as contingent. Not only do they
think that there are many possibilities concerning what specific bundles could have
obtained, they hold that it is possible that there are concrete particulars that are not
bundles of properties. This might be easier to understand if we resort to possible
worlds talk.
It is commonplace for neo-Humeans to express the great range of what they take
to be possible in terms of possible worlds: a possible world is a complete description
of a way the entire world could have been (if not the entity picked out by that
description). It captures a complete history of the universe (all the objects, properties,
relations, and so on) not just as it is, but as it would have been had this or that (or this
and that!) been different. In the manner of possible worlds speak, the neo-Humean
thinks that there are many possible worlds that match ours with regard to the basic
metaphysical facts, but claim that there are plenty of others that do not. They take the
basic metaphysical facts to be up for grabs just as much as the application of those
facts. There are worlds, they claim, where concrete particulars have no structure, and
others where they are built around substratum. There may even be worlds that have
all of the above, or one for some time and then others.¹⁷
I do not deny that the debate over the structure of concrete particulars—like many
other first-order metaphysical debates—is a live one, and a lively one, too. There is
some fact of the matter one way or the other, and it is fair to say that we do not know
what it is. But I do not see how our ignorance in these matters translates to genuine
ways the world could have been. If ours is a bundle theory world, then that is what it
is, knowledge be damned. We can—of course—imagine worlds where the basic
metaphysical facts differ from our own (an obvious truism; its being the case is an
antecedent condition of metaphysical debate). And we are unsure what sort of
metaphysic is active in the world. Throw in the neo-Humean’s liberal approach to
what is possible, and the neo-Humean arrives at the conclusion that we must treat the
alternatives as genuine possibilities.
I cannot help but think of these basic metaphysical facts as much more fixed than
the neo-Humean does. We can all make sense of what it would mean for these facts to

¹⁶ There is a larger set of metaphysical facts about the world that are not basic in this sense. They
concern individuals.
¹⁷ It may be that the neo-Humean has some incredibly basic metaphysical facts that hold in every world,
such as ‘any world must have a principle for the formation of concrete particulars’, but there are bound to
be some who deny even that.
   

differ, but why think we can make sense of them really being that way unless we are
just talking about what, for all we know, could have been the case? Our ignorance
should not be offered as evidence that something is possible. If the neo-Humean is
treating our expressions of ignorance as a basis for genuine possibility, then some-
thing has gone awry. Granted, our ignorance makes us unable to rule things out, but
that does not thereby rule them in. I am inclined to feel the same about conceivabil-
ity: how can our imagining something’s being the case constitute evidence for that
something’s being possible? This makes possibility awfully cheap. This is one of the
major pitfalls of this sort of thinking about possibility and possible worlds talk: if you
can describe a world, it lends artificial credence to thinking that things really could
have been that way.

2.3 Of Categoricalism
If you are comfortable believing that anything separable in one’s imagination is
similarly separable in nature, and thus attracted to the thought that there are no
necessary connections in nature, then you are likely to be okay with removing the
laws and causes typically associated with specific properties, and thinking that
properties—in and of themselves—have meagre essences. Laws and causes, you
would believe, are only contingently connected with those properties, and so the
real nature of the properties is what you find when they are stripped ‘bare’ of their
contingent causal profiles. But you need not stop there: everything modal has got to
go, too. Only when you remove all this contingent detritus do you finally get down to
the property’s real essence. Only then do you get to properties as the neo-Humean
understands them. Only then do you get categorical properties.
According to the neo-Humean, the most basic properties are categorical. To
describe a property as categorical is to say that it is in ‘pure act’; that is, exhibiting
all of its nature at all times. With categorical properties, what you see is what you get;
there is nothing more to the property that is hidden, and nothing it—unaided—is
liable to do.¹⁸ (By contrast, power properties keep much hidden, revealing parts of
themselves only in specific circumstances.) Categorical properties, in and of them-
selves, are thus entirely passive. They lack potential: they can only act if pushed
around by something outside them, like a law. According to the neo-Humeans, all
properties—all those that are owed a proper place in our ontology, anyway—are like
this. Given the ubiquity of the neo-Humean metaphysic, most readers should have
no trouble producing lists of putative examples of categorical properties. They are, we
are told, the familiar properties we ascribe to objects every day. They are such
properties as shape, size, number, colour, and so on.¹⁹
The neo-Humean conception of properties as categorical is clearly at odds with
that of properties as powers: whereas the latter understands properties as inherently

¹⁸ Well, sort of. You could not see a categorical property without causal intervention, and so observation
is no way to learn about the non-contingent nature of categorical properties.
¹⁹ As we will see, there is much dispute to be had here. Many—myself included—contend that all
properties are powers. Hence, it is part of the powers ontology that these properties will be powers, if they
are properties at all.
  -  

powerful, the former strips them of all such responsibility. A simple version of the
difference can be seen by contrasting the shape of a martini glass with its fragility. If
the distinction genuinely obtains, then its shape is not something that in and of itself
is full of threats and promises. Shape is just the topological arrangement of the glass’s
constituent molecules. Its fragility, however, is a capacity that is future-directed: it
threatens what sort of future will obtain if the glass is treated in the wrong way. It is
not, therefore, in pure act. With fragility what you see is not what you get.
You might find it strange to think that properties could fail to have potentialities
built into them. It is surely commonplace to think that a change in a thing’s
properties would mean a change in its capabilities. To wit, if you deflate a soccer
ball, it can no longer roll, and if you paint it black, it would be hard to find in an
unlit room. But there is an important distinction to be made between conceiving
of properties as having these sorts of potentialities built right into them versus
those potentialities being something that is added to them later. The distinction
concerns the essence of the properties: powers are essentially such that they have
their causal capacities built into them (properly understood, they just are those
causal capacities²⁰). By contrast, causal capacities play no part in the essence of
categorical properties. Categorical properties can acquire such potentialities in the
right environments, but in and of themselves they are inert. These acquired causal
capacities are not part of what a categorical property is.
Key to understanding properties as categorical requires seeing them as non-modal;
that is, not having any other ‘modes of being’ built into them. This is a rather odd way
of speaking, but all it means is that there is nothing more to the properties than what
they now display. Contrast this with a (putative) power like fragility: fragility has one
foot firmly planted in things as they presently are, and yet is directed at future states
that do not presently obtain (and might never obtain). Inasmuch as we understand
fragility in terms of a potential future state of brokenness, we are speaking of the
property in a qualified way, and so it is a modal property. To be directed at the future,
as any power property is, is to exhibit some other mode than that of being in pure act.
How, you might well ask, could anyone seriously maintain that all properties are
non-modal? After all, if fragility names a modal property, then excluding it from
what exists would demand one reject the obvious truism that glasses and a good
many other things are fragile. Surely this is nonsense. Unfortunately, the case against
the neo-Humean cannot be made so easily.
For starters, the neo-Humean need not deny that glass is fragile, nor that fragility is
a genuine property. Hers is the slightly weaker position that the sparse or funda-
mental properties are purely categorical. The sparse properties are all and only those
properties that carve nature at its joints (that is, they are those properties referred
to by our best scientific theories).²¹ They are fundamental to the extent that the
categorical properties are the most ontologically basic: all other properties rely on the
fundamental properties for their existence.²² On this view, the non-fundamental
properties are very much second-class existents, if they are existents at all. The

²⁰ Plus a little something more that can wait until Chapter 5. ²¹ Schaffer (2004).
²² See Sider (2012) and Hawthorne and Dorr (2013).
   

neo-Humean who countenances non-fundamental properties thus takes reality to

have a hierarchical structure. At its base are the properties the ontology cannot do
without. These are the properties the neo-Humean claims are categorical (though she
may claim the same of the non-fundamental properties, too).
When push comes to shove, the fundamental categoricals are the properties to
which the neo-Humean is most committed. As one moves up the structure, the
strength of commitment to higher-level properties diminishes. Some outright deny
that these second-class properties exist at all; others admit them, but to a lesser
degree. One way of picturing this concerns what actions a god would have to take to
create a world like ours. According to neo-Humeanism, a god need only create the
sparse set of categorical properties; the other properties come ‘for free’. A story about
the manner in which the base properties support them is required, but the god does
not lift another finger. By structuring things in this way, the neo-Humean is well
within her rights to speak of the glass’s fragility without violating her categoricalism
about properties. (It is worth noting that I, too, employ this hierarchical strategy, and
that I am primarily concerned with the most fundamental properties. At times—
typically to make examples more homey—I appeal to macroscopic properties to
illustrate features of the microscopic level, but my apparent commitment to those
macroscopic properties is a convenience. I will not yet take a stance on whether such
properties are second-class or worse.)
A key element of neo-Humeanism involves explaining how features the world
appears to contain, such as powers, can be ‘explained away’ by the categoricalist
framework. This sort of task is really the hard work of any metaphysic, and under-
lines the difference between an ontology and a metaphysic. An ontology is an answer
to the question what exists, but a ‘metaphysic’ tells you how instances of those
ontological entities give rise to the world as we know it. A metaphysic thus explains
how the things that seem to exist are accounted for via an ontology that does not
(strictly speaking) include them. The relevant part of the neo-Humean metaphysic is
the story of how objects have powers, couched only in terms of categorical properties.
The most common approach is the SCA (§1.3), which attempts to analyse the having
of a power (disposition) in terms of stronger-than-material conditionals, or ‘coun-
terfactual conditionals’. The story works in two steps: first, analyse the having of
powers without recourse to powers themselves, by appealing to counterfactuals, and
second, give an account of counterfactual conditionals that avoids commitment to
any properties that are not categorical.²³ We need not assess the success of that
reductive project here. The point is that powers seem to be a part of the world, so
those who deny there are any such properties owe us an account of how any would-
be powers are accounted for in an ontology that eschews them, which this attempts.
Another strategy available to the neo-Humean (one I also employ for non-
fundamental powers) makes use of the difference between predicates and properties.
It may well be true that the glass is fragile, but the truth of the corresponding
proposition does not demand that the glass possess the property of fragility, only

²³ It is bound to have other features, but it is not a worry so long as those other features are not powers
(or merely powers in disguise).
  -  

that the predicate truly applies. And all this demands is some truthmaker; that is,
some portion of the world that makes the proposition true.²⁴ For some predicates,
this will be a matter of the object in question possessing the corresponding property.
This is a common truthmaker for such claims, but there are true predications
whose truthmakers are more complicated, potentially involving the obtaining of
various states, facts, principles, properties, and arrangements thereof. Hence, it
might be that the glass counts as fragile in virtue of its having other properties
working in concert with one another in some specific nomological context, and not
fragility at all. (This sort of approach sits well with the austere neo-Humean ontology
mentioned above.)
This is part of a larger metaphysical lesson concerning the connection between
predicates and properties: the relationship between the predicates and properties is
not standardly isomorphic. There is sometimes a one-to-one correspondence
between a predicate we use and a property in the world, but it will just as often be
one-many, many-one, one-none, or none-one. Hence, it does not follow from an
instance of true predication that there is any one property the predicate picks out, or
any property at all. True predication requires no more of the world than that the
predication is apt. And that can be achieved by possessing a categorical property in
the right context, as imagined by the SCA.
By now you can hopefully appreciate that the neo-Humean has many strategies for
accommodating powers within her non-power ontology. To wit, there are plenty of
ways for the neo-Humean to make sense of the glass’s being fragile that do not
commit her to the sorts of properties a powers theorist endorses. If she did not, the
game would already be up. But it is far from up.
The nature of properties as categorical is the central aspect of the neo-Humean
metaphysic I seek to challenge. But the differences between the neo-Humean and the
powers theorist do not end there. An ontology and the metaphysic that surrounds it
are integrated in such a way that with a change in properties comes a change in how
we see many other aspects of the metaphysic. We have already seen how the neo-
Humean thinks differently about the nature of possibility, and that this ties in to her
conception of properties as categorical. I turn now to the two areas of the neo-
Humean metaphysic that the accounts of possibility and properties impact most
heavily: laws of nature and causation.

2.4 Of Laws of Nature

Much of the neo-Humean account of the laws of nature can be traced back to the
rejection of necessary connections in the world. This rejection undergirds the
thesis of Humean Supervenience, and fuels the following quintessentially neo-
Humean claims: (i) that the laws of nature are contingent; (ii) that the laws are
outside of, and distinct from, the objects they govern; (iii) that the objects subject
to those laws are inherently passive; and (iv) that causation lacks most or all of the

²⁴ See Smith et. al (1984) and Armstrong (2004a) for the basic idea, and Beebee and Dodd (2005) for
discussion.
     

oomph we typically think it does. The last I save for the next section; here I deal
with the first three.
For the neo-Humean, what is possible is guided (if not dictated) by what can be
conceived, and this leads to a rejection of necessary connections. The reasoning is
simple: imagine a case wherein two things—events, say—are putatively necessarily
connected. This putative necessary connection might be of the form that the occur-
rence of the first necessitates the occurrence of the second, in the way causation is
commonly understood. But now imagine one without the other. We can, and
therefore the neo-Humean believes the putative necessary connection between the
two is broken. Hence, no connection between them (causal or law-like) is one in
which the former event necessitates the latter. So, neither laws nor causes necessitate
their effects. And because we can further imagine there having been different laws or
none at all, the laws are contingent in a second sense. The laws neither act nor hold of
necessity. This is a brief characterization of the neo-Humean’s negative account of
the laws of nature. For the neo-Humean’s positive account, we need to build from the
bottom up. And that means starting with the doctrine of Humean Supervenience.
When presenting his doctrine of Humean Supervenience, Lewis employs a meta-
phor of the world as a ‘vast mosaic’.²⁵ The mosaic metaphor is a good tool for
understanding supervenience and the neo-Humean account of the laws of nature.
A mosaic is a piece of art constructed out of ever-so-many small tiles of coloured
ceramics or glass, typically arranged so as to create some image or pattern. The
patterns themselves are not found in any tile in particular; they only appear or
emerge from groups of tiles once they are arranged. The pattern is thus said to
‘supervene’ on the arrangement of the tiles; the arrangement of the tiles determines
what pattern emerges. It follows that one could not have a different pattern without
moving the tiles.²⁶
In typical cases, the patterns or images are formed by following a plan: the tiles are
placed in order with the resultant image in mind. But what happens if the tiles are
arranged without a plan? Rather than following any rules or guidelines about tile
placement, imagine the tiles are randomly distributed throughout the mosaic. The
distribution is random, but patterns nevertheless emerge. Not intended patterns, nor
patterns the artist sought to create through a specific ordering of tiles, but patterns
nonetheless. If one were to track those patterns through the mosaic, one could begin
to develop rules that would appear to have governed the placement of tiles. But they
would only appear to have done so; we know that no plan was followed. (This is the
equivalent of tea-leaf reading, and like tasseography falsely suggests plans where
none exist.) However, if these after-the-fact rules are carefully developed, using the
largest possible region of the mosaic, then they will serve as an excellent guide for
navigating the mosaic, despite not having served as a plan for its construction. Faced
with an infinitely large mosaic, these after-the-fact rules would prove very handy
indeed, rather like a map does, despite not having determined the landscape.
With this idea of after-the-fact rules at hand, shift now to properties. We find
properties littered about the universe. Think of this spatiotemporal distribution of

²⁵ Lewis (1986b: ix). ²⁶ Kim (1982, 1990, 1993a), McLaughlin (1995), and Jackson (1998).
  -  

properties (not just the present distribution but past and future as well) as tiles in a
huge mosaic. Just like the artistic case, patterns emerge, as do after-the-fact rules
concerning the relative locations and orderings of sets of properties in the universe.
These after-the-fact rules are the neo-Humean’s laws of nature. They do not govern
in the sense that they require (even weakly) that this or that property come after or
before some other, or be partnered in this or that way; they do not determine the
arrangement of the properties or the pattern that emerges. But from the patterns that
randomly arise we can develop useful guides to the universe and its navigation—
guides that the neo-Humean understands as the laws of nature. The archetype of this
account of laws is the Mill–Ramsey–Lewis ‘best systems’ account, according to which
laws are those generalizations that best systematize this information, where ‘best’ is a
balance of simplicity and strength.²⁷
These laws are after-the-fact laws. They do not push around the objects or their
properties; they do not govern. But nor do the objects or their properties propagate
their own movement and change (as the powers theorist claims). The neo-Humean
metaphysic thus sees physical objects as inherently passive: objects are lifeless
puppets whose apparently perfectly choreographed dance is really just a series of
random positions. The dance is not choreographed at all; it is a log book, and nothing
more. Objects appear in various locations and configurations—their properties
sometimes changing, other times not—but nothing makes this the case; all changes
and similarities are ‘explained’ after the fact. There is no saying why such and such
occurred, only that it did, and that it occupies a place in a wider pattern of such
happenings. If you want to know why something happened, the neo-Humean refers
you to the after-the-fact laws. Natural laws are thus stripped of their causal efficacy
and demoted to the status of mere statements and facts about the positions and
formations the objects and their properties happen to occupy.
Understood this way, laws of nature are powerless, but clearly far from useless.
There is little more we could want from a mature science than what the strongest and
simplest map of the distribution of properties could offer. In fact, in some respects
I have made this view sound worse than it is. One thing to be said in its favour is that
it coheres with basic operating procedure in the natural sciences. For those sciences
that offer laws, those laws are the most unified and systematic generalizations of the
patterns of property distributions we find in the world. They track behaviours and
patterns, and they predict on that basis. Our application of laws within science
demands little more of our metaphysics than the neo-Humean picture delivers.
But one should not mistake scientific practicality for metaphysical fact. It is the
role of science to track patterns and positions, and to make predictions on that basis.
It is not for the scientist to answer what ultimately affords these positions and
patterns. Moreover, if you are inclined to think that laws govern what occurs—that
is, if you have the common intuition that laws dictate the behaviours of objects and
form part of the explanation of why the properties are distributed as they are—then
this picture of laws will not ultimately be for you. It offends against the sense that the

²⁷ Mill (1967), Lewis (1973b, 1983b, 1999), Ramsey (1978), Beebee (2000), and Cohen and Callender
(2009).
     

world has more oomph to it than after-the-fact laws provide. The natural sciences
do not demand that the world be governed by the laws we locate in the patterns,
but we might want to insist that there is oomph nonetheless. And our desire for
oomph would lead us to the conclusion that the after-the-fact laws are merely our
first step in learning what the real story is. It would follow that the neo-Humean
picture of laws falls short: it is sufficient for scientific practice, but not for the
scientists themselves.
For the purposes of contrast—especially as it pertains to laws—it is worth our
briefly considering the metaphysic that sits roughly between neo-Humeanism and
the powers metaphysic. This middle-ground position is that most commonly asso-
ciated with the work of David Armstrong. Like the neo-Humean’s, Armstrong’s
properties are categorical, and the laws are strictly independent of them. But unlike
neo-Humeanism, Armstrongian laws necessitate their effects: they actually govern.²⁸
These laws—call them ‘strong’ laws—match up with our pre-theoretical desiderata
regarding the laws of nature in that they push objects around in virtue of the
properties they possess. And though the properties and objects that bear them are
inherently inert, they take on a quasi-powered nature when in the presence of strong
laws. Furthermore, instances of those laws double as instances of causation, meaning
Armstrongian causation is both intrinsic and oomphy. On this view, then, the mosaic
is created according to a genuine set of rules, and is not random.
The Armstrongian metaphysic shares its property passivism with neo-Humeanism,
but stands apart through the robust modal element it adds in the shape of laws
and causes. And though these strong laws are contingent (they could have been
different), they nevertheless impact the Armstrongian’s space of possibilities.²⁹ The
Armstrongian metaphysic is a clear departure from neo-Humeanism, but one has
to go further still to get to the powers metaphysic. Going further requires add-
itional changes in the nature of possibility, but, more importantly, it takes rejecting
passivism outright, by treating the properties as the seat of the laws. In other
words, you need to take the governing force of the Armstrongian laws and make it
part of a property’s essence. The force must be in the property, not outside it; it
must be incorporated into the very nature of property itself. Rejecting categorical-
ism in favour of an ontology of causal powers thus puts the rules that generate the
patterns of the mosaic into the properties themselves.
I find the strong account of laws far more attractive than its neo-Humean
counterpart, but both have their merits. As I say, it is not my aim to show that this
view or any like it is false; what I want to show is how the flavours of neo-Humeanism
blend together, so that we can properly assess what the powers metaphysic has to
offer without accidentally tipping the scales against it. We have tasted a good portion
of the flavour of the neo-Humean metaphysic, with regards to possibility, properties,
and laws. Last up is causation.

²⁸ Armstrong’s account of laws is known as the ‘DTA’ account, because it was independently proposed
by Dretske (1977), Tooley (1977), and Armstrong (1978, 1983).
²⁹ Meaning that the Armstrongian metaphysic differs from neo-Humeanism on three of the four main
components listed at the end of §2.1.
  -  

2.5 Of Causation
The final cog in the neo-Humean metaphysic concerns causation. It should come as
little surprise that the neo-Humean account of causation is decidedly weak, in the
sense that causal relations pack no oomph. On this picture of causation, causes do
not push the world towards effects as much as they just stand next to them. This is
well and truly Hume’s legacy. Nevertheless, despite rendering causes impotent, the
neo-Humean account of causation is an attractive and useful account—at least to the
extent that we want an easy-to-understand account of causation that plays well with
our everyday folk concept. I ultimately believe that the best answer to the question of
what causation is depends on the exercising of powers, but that makes the neo-
Humean account no less useful for other purposes.
To build up to the neo-Humean account of causation it will be handy to start with
a short refresher on Hume’s account. If the popularized (perhaps cartoonish) version
of Hume represents him as famous for anything, it is as being a great sceptic when it
comes to causation. Some historians of philosophy have questioned this reading, but
my purpose here is with flavours and not facts, and so we can leave the history to the
historians and make do with the CliffsNotes version of Hume.³⁰
So, what did Hume say about causation? The short version is that he examines the
world looking to find the source of his idea that causes necessitate their effects, and
comes up short. He concludes that the powers or forces we credit causes as possessing
have nothing of the sort: we add this in our understanding of them. What he claims is
in fact present, and thus what forms the basis of his positive account, are the
following: temporal succession (the one event precedes the other, often immediately),
and spatial contiguity (the two events occur in the same space). But as Hume is well
aware, this cannot be sufficient for causation, as not all pairs of events related in this
way are instances of causation. Something more is needed. Hume’s response is that
we move away from the specific case to something we find in similar cases, something
we find in all similar cases. That is, we move to a regularity. When we consider the
single case, and/or we see the cause event, we recall the many times we have seen a
(causal) pair of events, and infer that the effect will follow. We are thus conditioned
to infer the effect. This gives rise to the third feature, that of constant conjunction:
we consistently (perhaps unfailingly) find the cause paired with the effect. We come
to ‘see’ them as connected. He thus arrives at the following account of causation: an
event c is a cause of an event e just in case: (i) c is followed by e; (ii) c and e are in the
same region; (iii) c and e are events of kind K₁ and kind K₂ respectively, where
events of kind K₁ are succeeded by events of kind K₂. This is known as a ‘regulatory’
view of causation.
You might already have noticed the parallels between Hume’s account of causation
and the neo-Humean account of laws. Both give a central place to pattern-tracking.
In one case those patterns ground the laws, in the other they constitute the basis
on which pairs of events are causally connected. It follows that the Humean account
of causation is extrinsic: what makes it the case that a pair of events is causally
connected depends on there being patterns of exactly similar pairings of events of

³⁰ For recent and non-canonical interpretations of Hume, see Strawson (1989) and Winkler (1991).
   

that type, elsewhere and elsewhen. Moreover, intrinsically causally paired events
constitute a violation of Hume’s Dictum, and so extrinsicality is the only game in
town. On the traditional Humean view, the causal connection is also subjective (the
connection is something we draw from our experience), but contemporary regulatory
views of causation drop this aspect, bringing them more naturally in line with the
neo-Humean account of natural laws.
The classic Humean regulatory account of causation meets with some similarly
classic objections.³¹ The most pressing of these concerns the possibility of spurious
causation. As presented, the regulatory account of causation would count any regular
pairing of events as causal, even if we took this to be merely accidental. As a simple
case, imagine that a dog barks all and only those times that a specific light switch is
flicked to the up position. The account would mistakenly imply that the flicking is the
(or a) cause of the barking. Similar cases are easy to come by, and there is no obvious
way for the Humean to rule them out without importing assumptions that run
counter to the account.³² Let us put aside the extent to which Hume or the Humean
can satisfactorily respond, so I can highlight what I find important about the
objection. What it suggests is that we think there is something to genuine causation
the account fails to capture. That we can sort regularly paired events into those we
think are mere accidents versus those we think belong together—prior to considering
any account of causation—suggests that we take causal relations to be intrinsic. We
can, of course, be wrong about causes (maybe the switch really is causing the dog to
bark), but we see the pattern as the expression of the causal facts, and not constitutive
thereof.
There is more the Humean can say in support of the regulatory view, and more
that can be raised in objection to it, but what I have said is enough for our purposes.
What I want to do now is move on to the neo-Humean account of causation. The
neo-Humean account of causation I have in mind is the one popularized by Lewis in
the 1970s, which went on to create something of a philosophical cottage industry.
According to the Lewisian/neo-Humean account of causation, causation is to be
analysed in terms of counterfactual dependence. Like its forebears, it is an extrinsic
account of causation, in which causation lacks oomph. Hence it, too, runs foul of our
intuitions about the inherent force of the causal relation. However, unlike its prede-
cessor, it is nevertheless considered an attractive account of causation.
Inspired by a quote from Hume’s Enquiries, the counterfactual analysis of caus-
ation seeks to make good on Hume’s claim that we can understand a cause as
obtaining when some object is followed by another, ‘where, if the first object had
not been, the second never had existed’.³³ We have already met with one counter-
factual analysis, that which attempts to analyse what it is for an object to have a
power in terms of counterfactual conditionals. Here those same counterfactuals are
employed in an analysis of causation: causation is understood as a relation between

³¹ See Armstrong (1983) and Psillos (2002) for a few of them.

³² Not, that is, unless the Humean can relegate the differences to mere differences in how we treat the
cases. This is a level of subjectivism about causation that few would tolerate.
³³ Hume (1975: I.vii.29).
  -  

events where both obtain, but they are such that had the former event (the cause) not
occurred, then the latter event (the effect) would not have either.³⁴
As I have said, this is a reasonably attractive account of causation. Part of the
attraction is due to the account’s malleability: it applies just as well to microphysical
causes as it does to chemical reactions, human actions, historical events, and every-
thing in between. It is more or less a one-size-fits-all account of causation, and that is
certainly something to brag about. It also does a good job of capturing our everyday
use of the term ‘cause’, especially as an element of a sufficient condition for an effect.
In fact, as far as our common-or-garden concept of causation goes, you can do a lot
worse than the counterfactual analysis of causation.
Hold on—have I given up on powers already, just two chapters into a book-length
defence of them? Hardly; there is a distinction to be made between the best way of
understanding how we use the terms ‘cause’ or ‘causation’, and what, if anything,
counts as real causation in the world. I will have more to say about this distinction in
Chapter 6, but the lesson here is that the counterfactual analysis is well suited to the
first job, but it is powers that do the latter one best. I think the concept versus genuine
relation distinction also goes some distance to explaining why folks keep returning to
the counterfactual analysis (or close relatives) despite the massive literature on the
problems with the account (the still lively cottage industry previously mentioned).
Some of our concepts—like causation—are rough around the edges: they are not
constructed with well-defined boundaries or developed to deal with problem cases
like symmetrical overdetermination or trumping.³⁵ The analysis thus has a good fit.
But when folks raise problems like these for the counterfactual account, they dem-
onstrate the difficulties it has when understood as capturing genuine causal relations
in the world. As an account of what causation in the world really is, that is, what
makes the world tick, I think the counterfactual analysis fails on two counts. The first
concerns the many problems the account runs into, one of which I outline below. The
second reason is that it treats causation as weak and extrinsic. It thus lacks the oomph
that causation needs to have if it is to deserve the name.
Here is an example of the sort of problem the counterfactual analysis runs into
when seen as an account of how causation really operates. Imagine a cause with a
back-up mechanism, where the back-up mechanism makes sure the job gets done
even if the first should fail. For instance, you want to make sure that a bill gets paid,
and so you mail a cheque and submit payment online; or you shoot your mark in
addition to poisoning them; or you set two alarm clocks. Because the putative cause is
backed up by another would-be cause, the effect is not counterfactually dependent on
the original. Had the original not occurred, the back-up would have caused the effect.
Hence, there is no counterfactual dependence—and so no causal dependence—
between the putative cause and the effect. It is therefore not a cause. But of course

³⁴ Lewis (1973a).
³⁵ See Lewis (1986f), Schaffer (2000a, 2000b), and Collins et al. (2004). Symmetrical overdetermination
is when two processes lead to an effect for which each is sufficient. Trumping is a form of pre-emption in
which a sufficient process is cut off by another that ‘trumps’ it, as might happen when a policewoman and a
red light both demand you stop. (You would stop either way, but you obey the policewoman because her
order supersedes the red light.)
   

it is a cause, and so counterfactual dependence is not necessary for causation. So, even
though the counterfactual analysis gets many cases right, it runs into problems. That
is fine for our concept, but I have no idea what it would mean for real causal relations
in the world to ‘get it wrong’ sometimes.
Nevertheless, it is the extrinsicality that I find most problematic. On the counter-
factual analysis, causation holds between events. But events cannot themselves be
related by counterfactual conditionals, as those conditionals range over propositions.
So, the first step in the analysis is to replace every event with a corresponding
proposition. We then set up a dependence between families of propositions, of
which the event-capturing propositions are members. We next define causal depend-
ence in terms of the counterfactual dependence between the families of propositions,
such that an event e causally depends on an event c just in case the propositions that
correspond to e and c are members of counterfactually dependent proposition
families. That last step gets causation from causal dependence, as causal dependence
implies causation.
None of this yet explains why the account is extrinsic. The extrinsicality is a
by-product of the neo-Humean treatment of counterfactuals. The short story is
that the counterfactuals rely on the neo-Humean accounts of laws and of possi-
bility. In assessing the truth of the counterfactuals we ‘look’ to the goings-on of
nearby possible worlds. The worlds, recall, are complete descriptions of ways
things could have been; a nearby world is one that obeys the same (non-
governing) laws as the actual world, and which differs only minutely in terms of
the positions of the properties that make up the mosaic (it diverges just enough to
allow that the putative cause does not obtain, and then in all law-abiding ways
afterwards). It is through the ‘examination’ of these nearby worlds that we
establish what would have happened had the cause-event not taken place, and
how we determine if the effect-event depends on it.³⁶ The account is thus extrinsic
because it relies on extrinsic accounts of laws and possibility. We have to look
beyond the pair of events to determine whether they constitute a cause and its
effect. These same aspects render the account of causation weak: no event brings
about any other, they are merely related to each other according to patterns of
similarity and imaginability.
For those of us who think that causal connections are strong, that is, for those
of us who take causes to be oomphy in that they bring about their effects, the
counterfactual account will fall short of the mark. Like the regulatory account
before it, the counterfactual account boils down to sequences of events that lack
internal relations to one another. There is just one event, then another. Nothing
about the first demands, dictates, or makes the latter so. They just happen to be
found together. And where they just happen to be found together with great
frequency, we describe them as causally connected. This cannot be what caus-
ation is.

³⁶ I have put ‘look’ and ‘examine’ in scare quotes, because these are not genuine acts of looking or
examining; they are acts of imagining.
Another random document with
no related content on Scribd:
 So they waxed rich and happy, and there never was a time when a
man was hungry that he did not have some good things to eat, and it
very seldom happened that any of these hard workers found himself
without an appetite at meal-time.
For people who work hard and well are very apt to have all they
want and to want all they have. If they do not want it to use
themselves, they want it to sell or give away.
So, in time the people of this country became not only very
comfortable but very wealthy.
They had great barns full of grain and vast stores of everything
needful for their use and livelihood, and as they often sold their
surplus productions to other nations, they had great vaults full of
money.
But they all worked away every day, just the same as they used to,
because they were so accustomed to toil, that they would not have
been happy without work.
So, of course, they became richer and richer, and jollier and jollier
until at last they became so prosperous and happy that other nations
began to take notice of them. It was rather unusual, in those days to
see a whole nation so jolly.
The people in the adjoining countries were by no means so happy
and prosperous. Most of them were much better pleased with
fighting than with work, and it, therefore, often happened that they
were hungry when there was very little to eat.
For war is a very bad thing for crops. It is sometimes as injurious
as a long drought. For somebody must plant and hoe or there will be
little to eat in a land, and if the people spend most of their time in
warfare there cannot be much agricultural work going on.
But these outside people, especially those who lived in the land of
Voldor to the north of the country of the Cabordmen, had an idea that
it was a great deal easier to make war and capture supplies than to
raise crops themselves.
 This is why, after having carefully watched the Cabordmen for
some years, and noting their great possessions, they resolved to
make war upon these industrious and jolly people.
So they gathered together an army, which was an easy thing for
them to do, and invaded the country of the Cabordmen.
Our jolly friends were much astounded and distressed when the
great army of the Voldorites marched over their borders.

THE VOLDORITES MARCH INTO THE COUNTRY OF THE

CABORDMEN.
Now the poor Cabordmen knew not what to do. They were not
soldiers, and, indeed, there was not so much as a single sword or
spear or shield in the whole country. They never had gone to war
and they were not prepared for it, nor did they know anything about
fighting. It was altogether a new business to them.
 They gathered together and held hasty consultations, but they
could decide upon no plan to repel the invaders. What could they—a
nation of simple, jolly husbandmen—do against a great army of well
armed and practised warriors?
There seemed to be nothing left for them but to surrender at once,
and let the Voldorites help themselves to whatever they wanted. In
this case the poor Cabordmen and their families would not only be
stripped of every thing, but it was very likely indeed that the invaders
would carry off many of them as prisoners, and take them to Voldor,
and make them cultivate the land of their captors.
This was terrible to think of. But they could devise no plan to
escape this dreadful fate.
The Voldorites were now encamped upon the northern edge of
their territory, which was yet uninhabited and barren. The enemy so
far had met with none of the Cabordmen, but many of the latter had
seen the great army from afar without having made themselves
visible.
Night came on while the people were in this fearful condition of
fear and suspense. Less than a day’s march would bring the fierce
enemy into their midst. No one went to bed, for who could sleep at
such a time? No fires or lamps were lighted. They all gathered
together by the faint light of the new moon, and bewailed their sad
condition.
There was only one person among them who seemed to have
retained his courage and thoughtfulness. This was a young man
named Adar Gan Ip.
He was named Adar because he was a painter. Ip was his family
name, and he was called Gan after his grandfather. He was the only
painter in the whole nation, and he had learned his trade in a
neighboring country, where he had been to sell grain.
He principally painted signs and portraits. He did not paint many
portraits, because the people had but little time to sit for them, but he
painted a good many signs on barns and granaries. People liked to
have their names on their barns. He had no paint but one pot of
white paint. So when he painted portraits he painted only old men,
so that the white paint would do for their hair and beards as well as
for their faces. Having no colored paint for eyes, he always painted
portraits with the faces turned around, so that the eyes could not be
seen.
This young man was, as I have said, the only person among the
Cabordmen who seemed to have his wits about him.
He conceived a plan of safety, and lost no time in putting it in
execution.
The Cabordmen placed great confidence in him because of the
excellence of his portraits, and so when he told them his plan—or
that part of it which they were to carry out—they agreed to it at once.
What they were to do was very simple; each person was to take
two days’ provision, and to clear out of the country, every man,
woman, and child of them. They were to march away as fast as they
could over the south border, and to stay there until they heard from
Adar Ip. They were to take nothing with them but their two days’
provision and the clothes they wore, which were generally scanty, as
the climate was mild, and were to leave their houses and fields, and
everything just as they were at that time. Doors all open, and
everything lying where it had been last used.
So up got every man, woman and child, took food for two days,
and departed, leaving Adar Ip behind. They were all great walkers,
being so accustomed to activity in the field, and before morning they
had all passed out of sight over the south border of the land.
Then with his pot of white paint in one hand, and his brush in the
other, went Adar Ip, at the first peep of day, to the grave-yards of the
Cabordmen. There were three of these, not very far from the centre
of their country, which was a small country as you may well imagine.
The Cabordmen, being very healthy, seldom died of any disease
but old age; and there were not very many persons buried in the
three grave yards. In the first, and largest, there were seventy-two
graves; in the second, forty-one, and the third, a new one, only
thirteen. The graves were all leveled and sodded over, so that the
surface of the grave yard seemed like a beautiful lawn.
In one enclosure were the grandfathers, in another the
grandmothers, and in the third the very old maids and bachelors who
had died. There were no grave-stones or anything of the kind, but at
the gate of each enclosure was a board, stating how many persons
were buried therein. Every time it was necessary, which was very
seldom, Adar Ip painted out the old number on the board and put in
a new one.
When our young painter reached the first grave yard he quickly
painted three ciphers after the figures on the board by the gate. Then
running to the second enclosure he painted a three and two ciphers
on that board, and on the third, he painted a six and a five and a four
after the figures that were already there. Then he hurried away and
hid himself.
In the course of the morning the Voldorite army reached the
settlements of the Cabordmen. They did not stop long at the first
houses, but hurried on, carefully looking out on every side for some
sign of resistance from the people. But they saw no such sign, and
they saw no people. This naturally surprised them very much. And
the farther they went the more they were surprised.
At last the leaders ordered a halt, and gathered together for
consultation.
“I cannot imagine,” said the chief, “what this means. We must look
out for some ambush or trap. By the way, has any one seen any of
these Cabordmen?”
Careful inquiries were made, but no one had seen a Cabordman
since they had entered the country,
“This is indeed remarkable,” said the chief of the Voldorites. “I
cannot imagine what it means. No ambush has been discovered, no
fortifications, no people. The houses are all open. Everything seems
as if no enemy were expected. All their valuables are here. Where
are they?”
 Nobody knew, but just then a man who had been in the vicinity of
the grave-yards came running to the place where the officers were
gathered together, and he urged them to come back with him and
see what he had seen.
They all followed him, and when they saw the boards at the
entrance of the enclosures they were utterly astounded.
“What!” cried the chief, walking from one enclosure to another,
“Here lie buried seventy-two thousand Cabordmen, and here forty-
one thousand and three hundred Cabordwomen, and here thirteen
thousand, six hundred and fifty-four unmarried Cabordmen and
women! Comrades, we have found them! The whole nation lies
buried here!”
A deep silence fell upon the group of officers, and upon the vast
body of soldiers that had gathered around them.
At length the chief spoke again:
“It must have been a terrible pestilence,” he said. “The whole
nation lies buried here. I have added up these figures. I know there
were not more than one hundred and twenty-six thousand nine
hundred and fifty-four of them all put together. They are all dead and
buried here. It must have been awful!”
Some of the officers and soldiers then began to whisper together.
Then some one said out loud that this must be a dreadfully
unhealthy country. Then some of them began to move away as if
they were going to the rear to attend to something important in that
direction. Then the chief mounted his horse and rode away, and in
ten minutes that whole army made up its mind that it would be
exceedingly imprudent to remain any longer in such an
unwholesome country, and away they all marched towards Voldor.
The farther they went the more frightened they became, and soon
a perfect panic pervaded the army, and they set off at the top of their
speed, horsemen and footmen for their own barren but salubrious
land.
 THE FLIGHT OF THE VOLDORITES.

Away they went over the hills and the plains, and in two hours
there was not a Voldorite in the land of the Cabordmen.
Then uprose Adar Ip, and fled towards the southern border to
inform his countrymen of their happy deliverance.
They all returned quickly and found everything as it had been left.
Nothing had been taken, for none of the invaders wanted anything
that had been in a land where such a terrible mortality had prevailed.
Great was the joy and great the gratitude exhibited towards the
ingenious young Ip. The people presented him with a well filled
granary, and ordered him to paint on its walls at the public expense,
the history of his exploit.
“I wonder,” said one old man, “who they thought buried all these
people, if everybody was dead.”
 “I don’t know,” said Adar Ip. “But I think that they had such a high
opinion of the industry and prudence of our people that they
supposed we had doubtless made suitable arrangements for a
contingency of this kind.”
After this, the Cabordmen were never again disturbed, and they
became jollier than ever.
 Transcriber’s Notes

pg 257 Changed: shelter of the Esquimax snow houses

to: shelter of the Esquimaux snow houses
pg 272 Changed: The small volcanes are more active
to: The small volcanoes are more active
*** END OF THE PROJECT GUTENBERG EBOOK TALES OUT OF
SCHOOL ***

Updated editions will replace the previous one—the old editions

will be renamed.

Creating the works from print editions not protected by U.S.

copyright law means that no one owns a United States copyright
in these works, so the Foundation (and you!) can copy and
distribute it in the United States without permission and without
paying copyright royalties. Special rules, set forth in the General
Terms of Use part of this license, apply to copying and
distributing Project Gutenberg™ electronic works to protect the
PROJECT GUTENBERG™ concept and trademark. Project
Gutenberg is a registered trademark, and may not be used if
you charge for an eBook, except by following the terms of the
trademark license, including paying royalties for use of the
Project Gutenberg trademark. If you do not charge anything for
copies of this eBook, complying with the trademark license is
very easy. You may use this eBook for nearly any purpose such
as creation of derivative works, reports, performances and
research. Project Gutenberg eBooks may be modified and
printed and given away—you may do practically ANYTHING in
the United States with eBooks not protected by U.S. copyright
law. Redistribution is subject to the trademark license, especially
commercial redistribution.

START: FULL LICENSE

THE FULL PROJECT GUTENBERG LICENSE
 PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK

To protect the Project Gutenberg™ mission of promoting the

free distribution of electronic works, by using or distributing this
work (or any other work associated in any way with the phrase
“Project Gutenberg”), you agree to comply with all the terms of
the Full Project Gutenberg™ License available with this file or
online at www.gutenberg.org/license.

Section 1. General Terms of Use and

Redistributing Project Gutenberg™
electronic works
1.A. By reading or using any part of this Project Gutenberg™
electronic work, you indicate that you have read, understand,
agree to and accept all the terms of this license and intellectual
property (trademark/copyright) agreement. If you do not agree to
abide by all the terms of this agreement, you must cease using
and return or destroy all copies of Project Gutenberg™
electronic works in your possession. If you paid a fee for
obtaining a copy of or access to a Project Gutenberg™
electronic work and you do not agree to be bound by the terms
of this agreement, you may obtain a refund from the person or
entity to whom you paid the fee as set forth in paragraph 1.E.8.

1.B. “Project Gutenberg” is a registered trademark. It may only

be used on or associated in any way with an electronic work by
people who agree to be bound by the terms of this agreement.
There are a few things that you can do with most Project
Gutenberg™ electronic works even without complying with the
full terms of this agreement. See paragraph 1.C below. There
are a lot of things you can do with Project Gutenberg™
electronic works if you follow the terms of this agreement and
help preserve free future access to Project Gutenberg™
electronic works. See paragraph 1.E below.
1.C. The Project Gutenberg Literary Archive Foundation (“the
Foundation” or PGLAF), owns a compilation copyright in the
collection of Project Gutenberg™ electronic works. Nearly all the
individual works in the collection are in the public domain in the
United States. If an individual work is unprotected by copyright
law in the United States and you are located in the United
States, we do not claim a right to prevent you from copying,
distributing, performing, displaying or creating derivative works
based on the work as long as all references to Project
Gutenberg are removed. Of course, we hope that you will
support the Project Gutenberg™ mission of promoting free
access to electronic works by freely sharing Project
Gutenberg™ works in compliance with the terms of this
agreement for keeping the Project Gutenberg™ name
associated with the work. You can easily comply with the terms
of this agreement by keeping this work in the same format with
its attached full Project Gutenberg™ License when you share it
without charge with others.

1.D. The copyright laws of the place where you are located also
govern what you can do with this work. Copyright laws in most
countries are in a constant state of change. If you are outside
the United States, check the laws of your country in addition to
the terms of this agreement before downloading, copying,
displaying, performing, distributing or creating derivative works
based on this work or any other Project Gutenberg™ work. The
Foundation makes no representations concerning the copyright
status of any work in any country other than the United States.

1.E. Unless you have removed all references to Project

Gutenberg:

1.E.1. The following sentence, with active links to, or other

immediate access to, the full Project Gutenberg™ License must
appear prominently whenever any copy of a Project
Gutenberg™ work (any work on which the phrase “Project
Gutenberg” appears, or with which the phrase “Project
Gutenberg” is associated) is accessed, displayed, performed,
viewed, copied or distributed:

This eBook is for the use of anyone anywhere in the United

States and most other parts of the world at no cost and with
almost no restrictions whatsoever. You may copy it, give it
away or re-use it under the terms of the Project Gutenberg
License included with this eBook or online at
www.gutenberg.org. If you are not located in the United
States, you will have to check the laws of the country where
you are located before using this eBook.

1.E.2. If an individual Project Gutenberg™ electronic work is

derived from texts not protected by U.S. copyright law (does not
contain a notice indicating that it is posted with permission of the
copyright holder), the work can be copied and distributed to
anyone in the United States without paying any fees or charges.
If you are redistributing or providing access to a work with the
phrase “Project Gutenberg” associated with or appearing on the
work, you must comply either with the requirements of
paragraphs 1.E.1 through 1.E.7 or obtain permission for the use
of the work and the Project Gutenberg™ trademark as set forth
in paragraphs 1.E.8 or 1.E.9.

1.E.3. If an individual Project Gutenberg™ electronic work is

posted with the permission of the copyright holder, your use and
distribution must comply with both paragraphs 1.E.1 through
1.E.7 and any additional terms imposed by the copyright holder.
Additional terms will be linked to the Project Gutenberg™
License for all works posted with the permission of the copyright
holder found at the beginning of this work.

1.E.4. Do not unlink or detach or remove the full Project

Gutenberg™ License terms from this work, or any files
containing a part of this work or any other work associated with
Project Gutenberg™.
 1.E.5. Do not copy, display, perform, distribute or redistribute
this electronic work, or any part of this electronic work, without
prominently displaying the sentence set forth in paragraph 1.E.1
with active links or immediate access to the full terms of the
Project Gutenberg™ License.

1.E.6. You may convert to and distribute this work in any binary,
compressed, marked up, nonproprietary or proprietary form,
including any word processing or hypertext form. However, if
you provide access to or distribute copies of a Project
Gutenberg™ work in a format other than “Plain Vanilla ASCII” or
other format used in the official version posted on the official
Project Gutenberg™ website (www.gutenberg.org), you must, at
no additional cost, fee or expense to the user, provide a copy, a
means of exporting a copy, or a means of obtaining a copy upon
request, of the work in its original “Plain Vanilla ASCII” or other
form. Any alternate format must include the full Project
Gutenberg™ License as specified in paragraph 1.E.1.

1.E.7. Do not charge a fee for access to, viewing, displaying,

performing, copying or distributing any Project Gutenberg™
works unless you comply with paragraph 1.E.8 or 1.E.9.

1.E.8. You may charge a reasonable fee for copies of or

providing access to or distributing Project Gutenberg™
electronic works provided that:

• You pay a royalty fee of 20% of the gross profits you derive from
the use of Project Gutenberg™ works calculated using the
method you already use to calculate your applicable taxes. The
fee is owed to the owner of the Project Gutenberg™ trademark,
but he has agreed to donate royalties under this paragraph to
the Project Gutenberg Literary Archive Foundation. Royalty
payments must be paid within 60 days following each date on
which you prepare (or are legally required to prepare) your
periodic tax returns. Royalty payments should be clearly marked
as such and sent to the Project Gutenberg Literary Archive
Foundation at the address specified in Section 4, “Information
 about donations to the Project Gutenberg Literary Archive
Foundation.”

• You provide a full refund of any money paid by a user who

notifies you in writing (or by e-mail) within 30 days of receipt that
s/he does not agree to the terms of the full Project Gutenberg™
License. You must require such a user to return or destroy all
copies of the works possessed in a physical medium and
discontinue all use of and all access to other copies of Project
Gutenberg™ works.

• You provide, in accordance with paragraph 1.F.3, a full refund of

any money paid for a work or a replacement copy, if a defect in
the electronic work is discovered and reported to you within 90
days of receipt of the work.

• You comply with all other terms of this agreement for free
distribution of Project Gutenberg™ works.

1.E.9. If you wish to charge a fee or distribute a Project

Gutenberg™ electronic work or group of works on different
terms than are set forth in this agreement, you must obtain
permission in writing from the Project Gutenberg Literary
Archive Foundation, the manager of the Project Gutenberg™
trademark. Contact the Foundation as set forth in Section 3
below.

1.F.

1.F.1. Project Gutenberg volunteers and employees expend

considerable effort to identify, do copyright research on,
transcribe and proofread works not protected by U.S. copyright
law in creating the Project Gutenberg™ collection. Despite
these efforts, Project Gutenberg™ electronic works, and the
medium on which they may be stored, may contain “Defects,”
such as, but not limited to, incomplete, inaccurate or corrupt
data, transcription errors, a copyright or other intellectual
property infringement, a defective or damaged disk or other
medium, a computer virus, or computer codes that damage or
cannot be read by your equipment.

1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES -

Except for the “Right of Replacement or Refund” described in
paragraph 1.F.3, the Project Gutenberg Literary Archive
Foundation, the owner of the Project Gutenberg™ trademark,
and any other party distributing a Project Gutenberg™ electronic
work under this agreement, disclaim all liability to you for
damages, costs and expenses, including legal fees. YOU
AGREE THAT YOU HAVE NO REMEDIES FOR NEGLIGENCE,
STRICT LIABILITY, BREACH OF WARRANTY OR BREACH
OF CONTRACT EXCEPT THOSE PROVIDED IN PARAGRAPH
1.F.3. YOU AGREE THAT THE FOUNDATION, THE
TRADEMARK OWNER, AND ANY DISTRIBUTOR UNDER
THIS AGREEMENT WILL NOT BE LIABLE TO YOU FOR
ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE
OR INCIDENTAL DAMAGES EVEN IF YOU GIVE NOTICE OF
THE POSSIBILITY OF SUCH DAMAGE.

1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If

you discover a defect in this electronic work within 90 days of
receiving it, you can receive a refund of the money (if any) you
paid for it by sending a written explanation to the person you
received the work from. If you received the work on a physical
medium, you must return the medium with your written
explanation. The person or entity that provided you with the
defective work may elect to provide a replacement copy in lieu
of a refund. If you received the work electronically, the person or
entity providing it to you may choose to give you a second
opportunity to receive the work electronically in lieu of a refund.
If the second copy is also defective, you may demand a refund
in writing without further opportunities to fix the problem.

1.F.4. Except for the limited right of replacement or refund set

forth in paragraph 1.F.3, this work is provided to you ‘AS-IS’,
WITH NO OTHER WARRANTIES OF ANY KIND, EXPRESS
OR IMPLIED, INCLUDING BUT NOT LIMITED TO
WARRANTIES OF MERCHANTABILITY OR FITNESS FOR
ANY PURPOSE.

1.F.5. Some states do not allow disclaimers of certain implied

warranties or the exclusion or limitation of certain types of
damages. If any disclaimer or limitation set forth in this
agreement violates the law of the state applicable to this
agreement, the agreement shall be interpreted to make the
maximum disclaimer or limitation permitted by the applicable
state law. The invalidity or unenforceability of any provision of
this agreement shall not void the remaining provisions.

1.F.6. INDEMNITY - You agree to indemnify and hold the

Foundation, the trademark owner, any agent or employee of the
Foundation, anyone providing copies of Project Gutenberg™
electronic works in accordance with this agreement, and any
volunteers associated with the production, promotion and
distribution of Project Gutenberg™ electronic works, harmless
from all liability, costs and expenses, including legal fees, that
arise directly or indirectly from any of the following which you do
or cause to occur: (a) distribution of this or any Project
Gutenberg™ work, (b) alteration, modification, or additions or
deletions to any Project Gutenberg™ work, and (c) any Defect
you cause.

Section 2. Information about the Mission of

Project Gutenberg™
Project Gutenberg™ is synonymous with the free distribution of
electronic works in formats readable by the widest variety of
computers including obsolete, old, middle-aged and new
computers. It exists because of the efforts of hundreds of
volunteers and donations from people in all walks of life.

Volunteers and financial support to provide volunteers with the

assistance they need are critical to reaching Project
Gutenberg™’s goals and ensuring that the Project Gutenberg™
collection will remain freely available for generations to come. In
2001, the Project Gutenberg Literary Archive Foundation was
created to provide a secure and permanent future for Project
Gutenberg™ and future generations. To learn more about the
Project Gutenberg Literary Archive Foundation and how your
efforts and donations can help, see Sections 3 and 4 and the
Foundation information page at www.gutenberg.org.

Section 3. Information about the Project

Gutenberg Literary Archive Foundation
The Project Gutenberg Literary Archive Foundation is a non-
profit 501(c)(3) educational corporation organized under the
laws of the state of Mississippi and granted tax exempt status by
the Internal Revenue Service. The Foundation’s EIN or federal
tax identification number is 64-6221541. Contributions to the
Project Gutenberg Literary Archive Foundation are tax
deductible to the full extent permitted by U.S. federal laws and
your state’s laws.

The Foundation’s business office is located at 809 North 1500

West, Salt Lake City, UT 84116, (801) 596-1887. Email contact
links and up to date contact information can be found at the
Foundation’s website and official page at
www.gutenberg.org/contact

Section 4. Information about Donations to

the Project Gutenberg Literary Archive
Foundation
Project Gutenberg™ depends upon and cannot survive without
widespread public support and donations to carry out its mission
of increasing the number of public domain and licensed works
that can be freely distributed in machine-readable form
accessible by the widest array of equipment including outdated
equipment. Many small donations ($1 to $5,000) are particularly
important to maintaining tax exempt status with the IRS.

The Foundation is committed to complying with the laws

regulating charities and charitable donations in all 50 states of
the United States. Compliance requirements are not uniform
and it takes a considerable effort, much paperwork and many
fees to meet and keep up with these requirements. We do not
solicit donations in locations where we have not received written
confirmation of compliance. To SEND DONATIONS or
determine the status of compliance for any particular state visit
www.gutenberg.org/donate.

While we cannot and do not solicit contributions from states

where we have not met the solicitation requirements, we know
of no prohibition against accepting unsolicited donations from
donors in such states who approach us with offers to donate.

International donations are gratefully accepted, but we cannot

make any statements concerning tax treatment of donations
received from outside the United States. U.S. laws alone swamp
our small staff.

Please check the Project Gutenberg web pages for current

donation methods and addresses. Donations are accepted in a
number of other ways including checks, online payments and
credit card donations. To donate, please visit:
www.gutenberg.org/donate.

Section 5. General Information About Project

Gutenberg™ electronic works
Professor Michael S. Hart was the originator of the Project
Gutenberg™ concept of a library of electronic works that could
be freely shared with anyone. For forty years, he produced and