Citation
Conceiving-of and Conceiving-That

Material Information

Title:
Conceiving-of and Conceiving-That
Creator:
Savage, Kevin
Place of Publication:
[Gainesville, Fla.]
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (81 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.A.)
Degree Grantor:
University of Florida
Degree Disciplines:
Philosophy
Committee Chair:
Ludwig, Kirk A.
Committee Members:
Witmer, D Eugene
Jubien, Michael
Graduation Date:
12/14/2007

Subjects

Subjects / Keywords:
Existence ( jstor )
Locution ( jstor )
Mental acts ( jstor )
Modal realism ( jstor )
Nouns ( jstor )
Phenomenalism ( jstor )
Physicalism ( jstor )
Reductio ad absurdum ( jstor )
Statues ( jstor )
Unicorns ( jstor )
Philosophy -- Dissertations, Academic -- UF
Genre:
Electronic Thesis or Dissertation
born-digital ( sobekcm )
Philosophy thesis, M.A.

Notes

Abstract:
Philosophers use technical notions of conceiving to argue for a number of theses. However, very little has been done to illuminate what this act of conceiving consists in. So, we may fall into one of several traps when relying on what we can conceive in order to argue for some theory or other. For one, we may fail to provide a valid argument or we may have assumed our conclusion in one of our founding premises. In order to ensure we are not falling into one of these traps we should gain a fuller understanding of what conceiving is. Once we do that we see that many things that have been assumed about conceiving are in fact not the case. In particular, philosophers have argued that our ability to conceive that particular things could have been the case shows us that minds are not physical. But this argument rests on the assumption that what we can conceive is always possible. But there is a type of conceiving, the performing of which does not allow us to draw conclusions about what is possible. That is because there is a type of conceiving which allows us to conceive impossible objects. And so, if we are not careful we may believe we are performing an act of conceiving relevant to demonstrating the possibility of what we conceive but turn out to be wrong in fact. Thus, our arguments which rely on what we can conceive would fail in those cases. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis:
Thesis (M.A.)--University of Florida, 2007.
Local:
Adviser: Ludwig, Kirk A.
Statement of Responsibility:
by Kevin Savage.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Savage, Kevin. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Classification:
LD1780 2007 ( lcc )

Downloads

This item has the following downloads:


Full Text






above as a confusion between conceiving of God and conceiving that God exist. There would not be any

problem because if one were confused in the way described he would not be both conceiving that God

exist and conceiving that God not exist. Rather, he would be conceiving of God and conceiving that God

not exist.


Supposing that God is not an impossible object (and thus exists), the confusion can be explained

away in terms of 'conceiving of' and 'conceiving that', but with a slightly different approach. On the

model above, what one does (if God exists) when he believes he has conceived that God does not exist

is conceive of a world without God. If God exists and is metaphysically necessary then a world without

God is an impossible object. So while one would be able to conceive of a world without God, he would

not be able to conceive that such a world exist.


Some might try to explain away the apparent conceiving that something impossible exist with

only the 'conceiving that' locution. And if that is possible, then one might wonder why we should admit

that there is a substantive difference between conceiving of and conceiving that. It looks like David

Chalmers already gave us the relevant distinction with his categories of prima facie and ideal

conceivability:


Prima Facie Conceivability: "S is prima facie conceivable for a subject when S is conceivable for

that subject on first appearances."


Ideal Conceivability: "S is ideally conceivable when S is conceivable on ideal rational

reflection." (Chalmers, 2002, p. 147)


One might think that conceiving of some object o is the same as prima facie conceiving that o exist and

that conceiving that o exist (on my account) is the same as ideally conceiving that o exist (on Chalmers'

account). But I hold that conceiving of o need not include anything about o's existence. Consider some










must first understand A,then it looks like we might be unable to perform an analysis of A. This would be

a problem for any conceptual analysis.


However, it is only a problem for analysis if we hold that the analysans must provide us with the

ability to grasp a new concept. That is not what I hold. Analysis gives us an explanation of a concept we

already grasp.


We do not gain the ability to grasp specific concepts through analysis. In order to be able to

analyze a concept, C, we must be already possess the concept or we would not be able to determine

when to apply C. If one argues to the contrary, he might as well argue that because we come to know

something if we correctly analyze knowledge the analysis of knowledge would thereby be circular. Not

at all: it simply means the analysis fits the state we are in with respect to it.


We cannot require that in order for one to analyze knowledge, he must not prejudice himself

toward some specific analysis of knowledge in virtue of what he already thinks knowledge is. Analysis

done correctly should provide us with an explicit explanation of what we already grasp. If we perform

some act that provides us with something we did not understand going in then either an analysis has not

been performed or something has been performed in addition to an analysis.


When we perform analysis of some concept, C, we consider candidates for application

conditions for C and judge which ones are application conditions for C. At least one of the ways we do

this is by conceiving of some object or situation, 0, and then deciding whether C applies to O. If we

decide that C does apply to O, then we move on to some other object or situation in our considerations

until we find one to which C does not apply. If C does not apply to O then we hypothesize which

characteristics) of O would need to be different for Cto apply to O.










be any obvious contradiction and it does not look like people who utter (p) must be irrational in uttering

(p) even if they hold that fictional characters are impossible beings:


(p) I have conceived of Sherlock Holmes but I am certain (and was certain at the time of

conceiving) that it is impossible for Holmes to exist.


But if we look at (q), there is some significant tension:


(q) I have conceived Sherlock Holmes to exist/that Sherlock Holmes be existent but I am certain

(and was certain at the time of conceiving) that it is impossible for Holmes to exist.


In order to derive a contradiction from (q) one would have to hold that conceivability entails possibility.

I do not here want to presume that conceivability entails possibility. But even if one does not hold that

all fictional characters are impossible there are reasons to think that at least some fictional characters

are impossible. An obvious example would be a round-square-shaped character. Even if we cannot

make sense of a round-square-shaped character, there are other types of fictional characters and

creatures we can make sense of that are nonetheless impossible for reasons other than being fictional.

In Naming and Necessity, Saul Kripke gave an argument that a unicorn is a type of fictional creature that

is impossible for reasons other than just its being fictional.9


Tiger is a species of animal. And unicorn is supposed to be a species in the same way that tiger

is. If we found some creature that looked superficially like a tiger but had different organs inside or

perhaps different DNA, then such a creature would not be a tiger. The same would go for unicorns. But

the difference between tigers and unicorns is that while we have some non-superficial ways of

identifying members of the species, tiger, we do not have any non-superficial ways of picking out

9Kripke, pp 156-158. Kripke discusses both the metaphysical and epistemological impossibility
of unicorns. I have here abridged and combined the metaphysical and epistemological
arguments into one.

















CHAPTER 1

INTRODUCTION


The fact that something is conceivable or inconceivable has been used as the basis for many

philosophical arguments. For example, the conceivability of various scenarios has been used to argue

for skeptical theses. The supposed inconceivability of an object being unconceived was used by

Berkeley to argue for phenomenalism. Contemporarily, the conceivability of so-called philosophical

zombies has been used to argue against physicalism.


The claim of the conceivability of any of the relevant situations relies on our having conceived

those situations. So, there is some act we must perform, the performing of which allows us to hold that

the situation is conceivable. Therefore, all of the aforementioned arguments, as well as others, rely on

our being able to perform this act of conceiving. But, if we do not understand what it is we are doing

when we conceive, we may not be able to recognize when we have successfully conceived. I will argue

in this study that there is a type of conceiving that has not been clearly recognized, but that we do

perform, namely, the type of conceiving is normally denoted with reports using the 'conceive of'

locution (and its various tenses and moods).


There are two complementizers 'conceive' ordinarily takes in assertions: 'of' and 'that'. It has

traditionally been assumed that whether 'conceive' takes the complement 'of' or 'that' is

inconsequential. Both 'conceive of' and 'conceive that' assertions report the same kind of cognitive

operation, according to the tradition. Further, it has been held that both what is impossible is not













LIST OF FIGURES
1 A pe n ro se triang le ........................................................................................... ........................ .... 7 6










So then what is it we are doing when we think we are conceiving particulars? This is where

Berkeley comes in. What he had right was that when think about objects we think about their

properties. Even if there is some underlying substratum, it is not about that that we think. When we

conceive about contingent particulars we need to leave open that the particulars instantiate properties

other than the ones they instantiate in fact or the ones we believe they do in fact.


There is no favored property by which we pick out ordinary particulars, but I am suggesting that

when we supposedly conceive particulars we are conceiving properties. More precisely, we are grasping

a concept or group of concepts. From there we can draw conclusions about what would be required for

something to fall under a specific group of concepts all at once. So we conceive of a round square by

grasping both round and square. We can then draw conclusions about what it would take for something

to be a round square.


Berkeley does use the 'conceive of' locution that I propose denotes a substantively different act

from the act denoted by 'conceive that'. But I want to suggest that Berkeley is confused as to the level

of generality that can be conceived. Berkeley argues that one cannot conceive of an object

unconceived. Suppose we have a tree named Joe. If one conceives of Joe, Joe cannot be unconceived.

And the same goes for any particular. But the mistake is in thinking that our conceivings can be about

Joe at all. What one can conceive of is some co-instantiation of properties that the tree normally

instantiates.20 And what that amounts to is grasping the concepts under which any tree falls. But if all

conceiving is at this level of generality, there is no particular that has been conceived at all. While one

may not be able to conceive of any particular and have that particular be unconceived, that is only

because particulars cannot be conceived at all. What we can conceive is that there be something or



20One can likewise not conceive that Joe be taller or shorter, etc. but rather conceive that there
be a co-instantiation of such-and-such properties.










concepts, then different uses would be particularly difficult to notice. If such a conflation is going on in

the argument that it is possible for philosophical zombies to exist, then the argument fails as it is not

formally valid. Instead of conforming to the form presented on page three, it would instead be of this

form:


1. (Vx)(Conceivablel(x) -> Possible(x))


2. Conceivablel(a) -> Possible(a) (1)


3. Conceivable2(a)


4. Possible(a) (2,3)


The problem then is that there is no valid rule of inference that will allow one to derive the conclusion,

4, from 2 and 3, i.e., the argument is not formally valid.


There is a further problem that we might encounter with an argument the success of which

depends on the use of an unanalyzed term. We may be in danger of begging the question if, unknown

to us, we have smuggled our conclusion into our premises, that is, if one of our premises is in effect the

conclusion worded differently. Consider, for example, the terms 'impossible' and 'inconceivable'.

Sometimes, when we say that something is inconceivable we mean it is virtually impossible or

impossible in the sense of being ruled out by what we know. If at least some senses of 'impossible' and

'inconceivable' are synonymous then it stands to reason that some senses of 'possible' and 'conceivable'

are synonymous. If one were to use 'possible' and 'conceivable' in the same sense in an argument like

Chalmers' then the argument would be of the following form:


1. (Vx)(Possible(x) -> Possible(x))










What I propose is that the problem lies in a confusion between conceiving of and conceiving

that. I suggested in the last section that one can conceive of impossible objects but not of the existence

of impossible objects. If one conceives that where the subject of the that-clause is an impossible object,

10, then the that-clause, taken alone, would entail the existence of 10. Likewise, if one conceives of IO's

existing then the content of the conceiving also entails IO's existing13 trivially. And it is the existence of

10 that is inconceivable, or so I have argued.


My proposal would get around the problem that results from Yablo's supposed counterexample

to Textbook Kripkeanism because, although there would be some act of conceiving performed when one

believes he is conceiving that God exists, it would not be a conceiving that God exists. For, there is a

confusion. But the confusion is not about the type of necessity that God possesses or some

misunderstanding regarding the other properties God is supposed to possess. Rather, the confusion is

about the type of conceiving one is performing. Although 'conceiving of' is the more natural phrasing

for conceiving reports, philosophers almost exclusively use the 'conceiving that' construction. This

preference prejudices us toward thinking that all conceiving must be of some propositional content.


If we do not presume that acts of conceiving must have propositional content, at least as

normally thought of, then we can explain away the apparent problem presented above. What is

impossible is that impossible objects exist. It follows (at least intuitively, if not logically) that it is

impossible that we can conceive that those impossible objects exist. This is because conceiving that

some objects exist requires specifying at least the relevant circumstances under which such an object

exists. But it is clear that no such conditions can be specified in the case of impossible objects. So, if

God is one of these impossible objects then it is inconceivable that God exists. But I have left it open

that one can conceive of God. And so, one can explain away the apparent problem in the argument


"This presumes an intelligible notion of property entailment.










conceiving that is different from the distinction between prima facie conceiving and ideal conceiving.

The difference is made evident by the fact that there are some things that can be conceived of (e.g.

round squares) that one cannot even prima facie conceive existing. Prima facie conceiving requires

some propositional content which further requires that one be able to conceive of conditions on the

existence under which those things the content is about exist. But object-types which fall under

concepts that are obviously conceptually incoherent cannot even be prima facie conceived. That is

because some concepts are such that grasping them immediately puts one in a position to recognize

that they cannot apply to any objects.


In section 3.3, I looked at a confusion one might have that would lead one to have a

phenomenalist view. The confusion is about the level of generality at which we can conceive. We

cannot conceive about specific particulars. What we can conceive about is types of particulars. There is

nothing that can be both conceived and unconceived. But that is only a problem for non-

phenomenalists if we conceive of particulars. But not only do we not have to conceive of specific

contingent spatio-temporal particulars, but we cannot do so. What we do when we purportedly

conceive of contingent spatio-temporal particulars is evaluate the conditions anything would have to

satisfy in order fall under some concept or concepts we associate with those particulars.


In chapter 4, I will further explain the act we report with 'conceive of' sentences. I will also

further explain the relationship between conceiving of and conceiving that. Lastly, I will look at how one

might use this distinction to argue that zombie arguments fail.










for example, believes that it is epistemically open that George exists. Consider the difference between

(I) and (m):


(I) I have conceived of Sherlock Holmes.


(m) I have conceived that Sherlock Holmes exists.


One need not think that it is epistemically open that Sherlock Holmes exists in order to sincerely and

truthfully utter (I) but the belief that it is epistemically open that Sherlock Holmes exists would be

required for one to sincerely and truthfully utter (m) as shown by (n) and (o):


(n) I have conceived of Sherlock Holmes but I am certain (and was certain at the time of

conceiving) that he does not exist.


(o) I have conceived that Sherlock Holmes exists but I am certain (and was certain at the

time of conceiving) that he does not exist.


It is evident that (n) is consistent while (o) is contradictory. Perhaps if the that-clause in (o) is put in the

subjunctive mood then the problem can be remedied:


(o') I have conceived Sherlock Holmes to exist/that Sherlock Holmes be existent but I am

certain (and was certain at the time of conceiving) that he does not exist.


There are theoretical reasons to be wary of (o'). First, (o') treats existence as a property. But existence

is not normally treated as a property. Second, even if we table concerns about treating existence as a

property, there is a further problem presented for 'conceive that' sentences which are about the

existence of fictional characters or types. Some philosophers hold that fictional characters are

impossible. Even so, those philosophers can sincerely utter (I). If we look at (p) there does not seem to













3.1 'Conceiving of' and 'Conceiving that'


Philosophers tend to be concerned only with the verb 'conceive' when it takes a sentential

complement, either a that-clause (conceiving that there is extraterrestrial life), a nominalized sentential

complement (conceiving there being extraterrestrial life), or an infinitive sentential complement

(conceiving there to be extraterrestrial life). They are concerned with the 'conceive that' construction in

part because there is a tendency in philosophy of mind to be concerned with propositional attitudes and

in part because of the interest in whether a proposition's conceivability entails its possibility.


I will not in this section directly address whether the content of our conceivings must be

propositional. Instead, I will consider whether the 'conceives that' construction is the only one that

should be considered. It is my view that we should also examine the 'conceive of' construction.


The reason that we should look at the 'conceive of' construction is that reports using the

'conceive of' and 'conceive that' constructions differ not only in the complements 'conceive' takes in

those reports but also in that they are reports of substantively different types of acts. If the two

constructions are intersubstitutable salva veritate in all reports then 'conceive of' and 'conceive that' are

likely synonymous and thus reports using the 'conceive of' construction would be reports of the same

type of act as that expressed in 'conceive that' constructions; the differences would be merely verbal

and not substantive. But, if they are not intersubstitutable salva veritate then there is prima facie

reason to think that reports with the 'conceive of' or 'conceive that' construction are reports of different

types of acts.


However, even if the two constructions are not intersubstitutable salva veritate, the differences

between the two constructions may still be merely verbal if one may say everything with the 'conceive










conceptual incoherence in p. There is undoubtedly more to conceiving-that but the aim of this section

was to give some of the minimal requirements on conceiving-that.


In section 4.3, I looked at the zombie-argument against physicalism. I argued that one way a

physicalist might object to the argument is by pointing out that one who purportedly conceives there to

be a world with zombies really is conceiving of zombies. If that confusion is committed by one who

alleges to have conceived that zombie-worlds exist then the way in which zombies have been conceived

does not provide evidence that it is possible for zombies to exist. And thus, the argument fails to be

valid. I further suggested that the way one might conceive of zombies is the way we can conceive of

impossible fictions and impossible pictures. We perform these acts by entertaining the appropriate

concept, be it simple or complex, and drawing some conclusions which do not have metaphysical

implications.










requires that one already has figured out the application conditions for all the relevant concepts need to

evaluate p. So, conceiving-of is a precondition on conceiving-that. Conceiving that p (where p is about

some x) further requires that one not detect that there are any incoherent concepts in the content of

conceiving of x.


In this section, I have discussed what conceiving-of and conceiving-that are. Further, I have also

discussed the differences between these two types of conceiving and the relation between them.


Conceiving-of is a two stage process whereby one first entertains a concept and then draws

conclusions regarding that concept's application conditions. The figuring out of application conditions is

modeled by the reduction ad absurdum. Unlike reduction ad absurdum, the aim of conceiving-of is not to

draw a contradiction, but rather to draw out whatever inferences one can about the application of the

concept, coherent or not.


Conceiving-that p requires first that one conceive of a, where p is about a. So, conceiving of a is

a precondition on conceiving that p. Conceiving that p further requires that one not detect and any

incoherence in the concepts under which x falls. So, while conceiving-of is a precondition of conceiving-

that, conceiving-that further requires one to evaluate a proposition. If one does not detect that the

proposition is necessarily false then one has conceived that p, either ideally or in a prima facie manner.25

If one does detect that the proposition is necessarily false, then one does not conceive that p.









251t is important to note that one will not be excluded from conceiving on the grounds that he
has a greater capacity for rational reflection. Even an ideal conceiver can prima facie
conceive that p where p is not ideally conceivable. Such a situation would be one in which
the conceiver merely abstains from conceiving to the extent his ability allows.










With the motivations and preliminary concerns taken care of, I will in chapter three move onto

the argument for the distinction between conceiving-of and conceiving-that. On the basis of this

distinction, I will also be arguing for two other theses. First, I will argue that we can conceive of

impossible things but not that those things exist. Second, I will argue that we cannot conceive about

particulars per se. Rather, we can only conceive particulars insofar as they fall under some type.










But there is a type of conceiving, the performing of which does not allow us to draw conclusions

about what is possible. That is because there is a type of conceiving which allows us to conceive

impossible objects. And so, if we are not careful we may believe we are performing an act of conceiving

relevant to demonstrating the possibility of what we conceive but turn out to be wrong in fact. Thus,

our arguments which rely on what we can conceive would fail in those cases.















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts

CONCEIVING-OF AND CONCEIVING-THAT

By

Kevin Savage

December 2007

Chair: Kirk Ludwig

Major: Philosophy



Philosophers use technical notions of conceiving to argue for a number of theses. However,

very little has been done to illuminate what this act of conceiving consists in. So, we may fall into one of

several traps when relying on what we can conceive in order to argue for some theory or other. For

one, we may fail to provide a valid argument or we may have assumed our conclusion in one of our

founding premises.


In order to ensure we are not falling into one of these traps we should gain a fuller

understanding of what conceiving is. Once we do that we see that many things that have been assumed

about conceiving are in fact not the case.


In particular, philosophers have argued that our ability to conceive that particular things could

have been the case shows us that minds are not physical. But this argument rests on the assumption

that what we can conceive is always possible.










have been different than we believe they are or how they are in fact. Anything that we believe picks out

or in fact picks out a particular uniquely could be other than how it is or how we believe it to be.


In chapter 2, I will discuss preliminary considerations for this study. Section 2.1 will discuss the

motivation for the project. In brief, if one either does not understand what conceiving is or does not

recognize all of the different ways which we conceive, confusions may result. The specific forms these

confusions can take will be looked at in 2.1. Section 2.2 will be an exploration of various prima facie

problems for the project.


In chapter 3, I will address the three assumptions of the third paragraph. In section 3.1, I argue

that there is a substantive distinction between conceiving-of and conceiving-that. From the observation

that 'conceive of' may take as its object an ordinary noun or noun phrase whereas 'conceive that' must

take a sentence as its object, I argue that the sentential objects of 'conceive of' may be nouns or nouns

phrases which have as their ersatz referents types of impossible objects. In section 3.2, I argue against

the suggestion that we may conceive that some impossible objects exist (or that there be impossible

objects). I suggest that the reason it seems that we can conceive that some impossible objects exist is

that we have confused conceiving-of for conceiving-that. In section 3.3, I suggest a way of

understanding Berkeley's argument for phenomenalism which suggests that Berkeley's mistake might be

based on a misunderstanding of the limits of what we can conceive. I argue that one could argue for

phenomenalism as Berkeley did if he were to think that we can conceive particulars. Further, I suggest

that this is a mistake and that what appears to be conceiving particulars is conceiving something

general. I argue that any time we think we have conceived a particular we have made a mistake

because we cannot conceive particulars.


In chapter 4, I further delineate the difference between conceiving-of and conceiving-that while

also showing how the two types of conceiving are related and explain why it is important to distinguish










If we were to analyze conceive in this way, then we would conceive of some act and ask whether

that conceived-of act is conceiving. If successful, then we would be conceiving of an act of conceiving.

Conceiving requires a conceiver. So, a successful analysis of conceive using this methodology would

require us to conceive that someone conceive; we must rely on an act of iterated conceiving. That

might then require us to first-personally experience some mind other than our own. But even

hypothetically we can only experience our own mind. So, we might be closed off from this method of

analysis.


Perhaps we can get around this worry if we merely acknowledge that the conceiver within the

iterated act of conceiving is the same as the conceiver in the first-order act of conceiving. So, in order

for one to evaluate what it takes to conceive that P, he must conceive that he conceives that P. But

what it is that I conceive when I conceive that I conceive that P other than conceiving that P is puzzling.

Regardless, whether or not I can perform iterated conceivings, iterated conceivings would not tell us

anything more about what conceivings are than first order acts of conceiving.


But then we do not have the same method of conceptual analysis that we normally do because

we would not be conceiving anything that is a candidate for something to which conceiving applies.

Rather, the mental act itself would be a candidate for conceiving.


I raise this worry not because it is particularly problematic but rather to highlight a difference in

performing a conceptual analysis of conceive from performing an analysis of some other types of

concepts. It is performed by looking at actual (and not hypothetical) acts of conceiving. They may be

framed as hypothetical acts, particularly when we consider what people are doing when they make

specific conceiving reports, but it is our own mental acts that we use as evidence.










Before I get to the discussion of conceiving-that and its relation to conceiving-of, I will discuss

what conceiving-of is, which has only been touched upon earlier. In the previous section, I suggested

that the reduction ad absurdum provides us with a model of conceiving-of.


If we examine what goes on in reduction we will get a better idea as to what conceiving-of

comes to. In a reduction we make an assumption in order to show that that assumption, either by itself

or along with other premises, entails a contradiction. So, in at least some cases, the assumption itself is

necessarily false and so impossible.23 And in some cases it takes quite a bit of work to reveal the

contradiction. That work which reveals the impossibility is analogous to conceiving-of. But even if we

do reveal that some premise is necessarily false we may draw entailments from it which are not

contradictions. Consider (RS):


(RS) Round(a) & Square(a)


Although (RS) is necessarily false we can draw inferences from (RS) which are not, such as (R) and (S):


(R) Round(a)


(S) Square(a)


So we can sincerely utter a sentence like "I have conceived of a round square and if there could be such

a thing it would be round." Conceiving-of is analogous to the mental act that first presents (RS) to us

and then gets us from (RS) to (R).


Of course, there are limits to the analogy between reduction ad absurdum and conceiving-of.

The aim of conceiving-of is not to draw a contradiction. Conceiving-of is an act we perform without



23There can, of course, be reduction within reduction. In some of those cases, there may be
assumptions made for purposes of reduction that are not impossible.










3.3 Confusions That Could Lead to a Berkeleyan View


In this section, I will consider the type of confusion that might lead one to the Berkeleyan

conclusion that nothing exists other than minds and mind-dependent ideas. Consider the following

well-known passage from Berkeley's Principles of Human Knowledge:


But say you, though the ideas themselves do not exist without the mind, yet there may be
things like them whereof they are copies or resemblances, which things exist without the
mind, in an unthinking substance. I answer, an idea can be like nothing but an idea; a
colour or figure can be like nothing but another colour or figure. If we look but ever so little into
our thoughts, we shall find it impossible for us to conceive a likeness except only between our
ideas. Again, I ask whether those supposed originals or external things, of which our ideas are
the pictures or representations, be themselves perceivable or no? If they are, then they are
ideas, and we have gained our point; but if you say they are not, I appeal to any one whether it
be sense, to assert a colour is like something which is invisible; hard or soft, like something
which is intangible; and so the rest. (Berkeley, 1.8)



We can only perceive objects by way of their sensible properties. The objects which instantiate sensible

properties are either themselves perceivable or are not. If they are, then there is nothing more to

objects than their sensible properties. If they are not, then one must suppose that there is some

intimate connection between objects and their sensible properties. That intimate connection is

supposed to be resemblance. In order for something to resemble a sensible property it must itself be

sensible. So, if objects resemble their sensible properties, objects must be nothing over-and-above their

sensible properties. So, there is nothing more to objects than how we perceive them. Without our

perceiving them, then they would not exist. Hence, supposed external objects are mind-dependent.


Berkeley assumes that the limits our ability to conceive is given by what we perceive: "Hence as

it is impossible for me to see or feel any thing without an actual sensation of that thing, so it is

impossible for me to conceive in my thoughts any sensible thing or object distinct from the sensation or










might be some other way to say what is said with (e) with another sentence using the 'conceive of'

construction.


(g) I have conceived of George being a minister.


On a natural reading, (e) seems to be about the utterer having some evidence that George is a minister

and also that the utterer cannot rule it out that George is a minister. On the other hand, (g) is not

obviously about some attitude toward George being a minister. In order to make clearer the difference

between (e) and (g), we can perform a test.


If what is conveyed with (e) is that it is epistemically open that George is a minister, then it

should be contradictory to assert both that one has conceived that George is a minister and that one

knows that George is not a minister. So we can test (e) and (g) with (e') and (g'), respectively:


(e') I have conceived that George is a minister but I am certain (and was certain at the time of

conceiving)7 that he isn't (and wasn't) a minister.


(g') I have conceived of George being a minister but I am certain (and and was certain at the

time of conceiving) that he isn't (and wasn't) a minister.


It is clear that (g') is not contradictory, while (e') only seems non-contradictory if one forces an unnatural

reading.


One might attempt to alter (e) in some other way than in (e') in order to get the result that what

is expressed by (e) is also expressible with some sentence using the 'conceive that' construction:




7The parenthetical clause here is included to guard against a reading to the effect that the
knowledge that George isn't a minister came after conceiving that George is a minister. On
that reading, the test would be invalid.










CONCEIVING-OF AND CONCEIVING-THAT


By




KEVIN SAVAGE


















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS




UNIVERSITY OF FLORIDA




2007










states.


4. This world has mental states.


5. It is conceivable that there be a minimal physical duplicate of our world which is not a duplicate

simpliciter of our world. (3, 4)


6. If it is conceivable that there be a minimal physical duplicate of our world which is not a

duplicate simpliciter of our world then it is possible that there be a minimal physical duplicate of

our world which is not a duplicate simpliciter of our world. (Thesis that conceivability entails

possibility)


7. It is possible that there be a minimal physical duplicate of our world which is not a duplicate

simpliciter of our world. (5, 6)


8. There is some world which is a minimal physical duplicate of our world which is not a duplicate

simpliciter of our world. (Restatement of 7 through the use of possible worlds semantics)


9. Physicalism is false. (1, 8)


One might worry that this talk of possible worlds takes them too seriously. I am only using

possible worlds here for heuristic purposes and remain neutral on the matter of whether or not they

exist. The argument above may also be put in non-possible worlds terms:


Let P express a proposition that exhausts the physical facts of the world and let M express a

proposition that exhausts the mental facts of the world. If physicalism is true, then P entails M.


1. It is conceivable that there be zombies, i.e. the P be true and M be false.


2. If it is conceivable that P be true and M be false then it is possible that P be true and M be false.










when as a matter of fact, E-worlds cannot exist. So it is, for instance, with the conceivability of
water in the absence of hydrogen, or Hesperus without Phosphorus.

The good news is that (although conceivability evidence is fallible) the failures always take a
certain form. A thinker who (mistakenly) conceives E as possible is correctly registering the
possibility of something and mistaking the possibility of that for the possibility of E. There are
illusions of possibility, if you like, but no outright delusions or hallucinations.1




Textbook Kripkeanism assumes something about the connection between conceivability and

possibility. But one need not assume that conceivability entails possibility to hold onto the spirit of

Textbook Kripkeanism. I believe the spirit is captured by MC.


Yablo's examination of Textbook Kripkeanism is particularly relevant to the present issue

because Yablo contends that one may conceive of impossibilities.


Yablo's counterexample to Textbook Kripkeanism centers on the conceivability of a necessary

being. The argument goes like this (for some S):


1. Conceiver S can conceive that God exists.


2. S can conceive that God does not exist.


3. If S can conceive that P then P is possible.


4. So, it is possible that God exists and also possible that God does not exist. (1, 2, 3)


5. God exists in some possible worlds and does not exist in other possible worlds. (analysis of 4

using the heuristic of possible worlds)


6. But, if God exists, he is a necessary existent.


1 Yablo, Stephen. "Textbook Kripkeanism and the Open Texture of Concepts," Pacific Philosophical Quarterly 81
(2000): p. 108










(h) I have conceived that George be a minister.


If we perform the test with (h) that we did with (e) then we get:


(h') I have conceived that George be a minister but I am certain (and was certain at the time of

conceiving) that he isn't.


(h') is not contradictory so (h) does not convey that it is epistemically open that George is a minister.

Rather, (h) seems to be about one's consideration of a hypothetical situation. Further, the utterer of (h)

is reporting that he entertained a hypothetical situation in which George was a minister. Whether or

not the act reported by (h) is precisely, or only, the consideration of a hypothetical situation is not at

issue. The important point here is that (e) and (h) are not reports of the same act. Also, when (h) is

compared to (g), it is difficult to see a difference in acts reported by the two sentences. It looks like (g)

and (h) report the same act. If all 'conceive of' sentences can be used interchangeably with some

'conceive that' sentence as with (g) and (h) then there may not be any substantive difference in the acts

reported with sentences using each of the phrases.


But so far what has been suggested by the examples is that what is expressed with all 'conceive

of' statements can also be expressed with some 'conceive that' statement'. I want to suggest that there

is some act reported by at least some 'conceive of' statements that is substantively different from acts

reported by 'conceive that' statements. It is no matter to this study that it may not be that there are

acts reported with 'conceive that' statements that may not be reportable by any 'conceive of'

statement. If that is true, then it would be consistent with the position that acts reported with 'conceive

of' are merely a subset of the set of acts reported by 'conceive that'. That would not give evidence that

there are two substantively different acts, but merely some that cannot be reported with one of the

locutions for what might be merely pragmatic reasons.










whose DNA sequence is also explicitly stated in the story. However, the DNA sequence along with the

physical laws entail that Callie be brown-haired. Even if we are aware of this problem, we do not have

any problem conceiving of Callie and her blond-hair. One might argue that what we do in such an

instance is really think of the story as having different physical laws or Callie's DNA being different. But

we need not do so. All we need do is ignore the impossibility as in a reduction ad absurdum before we

conclude that there is an impossibility.


I suggest that we conceive of zombies in the same way we might conceive of Callie. Further, I

want to suggest that the difference between the appearance that zombies are conceivable and zombies

in fact being conceivable is not merely a matter of the degree of rational reflection, but rather a

confusion between types of conceiving.


Consider the penrose triangle in figure 1. A picture of a penrose triangle represents surfaces

that go away from and come toward a viewer at the same time. A true penrose triangle is impossible,

but we can look at the picture and understand what it depicts. And we can look at the picture all at

once. But it is not until one considers the picture as representing a way that things could be that a

problem arises. When one looks at the picture merely as a picture there is no problem. By this I mean

that some impossible things may be represented with pictures and we do not have any trouble

understanding the picture as such. An inconsistent co-instantiation of properties may be represented.

And, so long as we do not look at the picture as something to be evaluated we do not have any problem.

Of course, these pictures are more interesting after evaluating them as pictures of impossible things.

But, no matter, we can look at them without evaluating them and understand what they depict.


I want to suggest that that is what goes on with zombie-conceivings. We may consider what is

in some sense a representation of a zombie world. And we even grasp all of the concepts and group

them together as one. And in this way we can conceive of zombie-worlds just as we can look at pictures










perception of it" (Berkeley, I. 5).14 It is impossible to conceive of a sensible object without sensible

properties. Clearly, what it is to be a sensible object is to be something which can be perceived.


But Berkeley is here confused. The confusion commonly attributed here to Berkeley is that he is

confused between the objects of perception and the objects of introspection.


I do not want to suggest that Berkeley is not confused in the way people normally suggest he is.

What I want to do here is suggest that there is a further mistake Berkeley may be making. The mistake

lies in thinking that one can conceive of particulars as such. We normally do act as though conceiving

can be about particulars as such. But I want to suggest that that is something we cannot do (some

exceptions to be mentioned aside).


First, I will examine how holding that our conceiving can be about ordinary particulars could lead

one to phenomenalism. I will then argue that acts of conceiving cannot be about ordinary contingent

particulars.15 We should look at an example in which we purportedly do conceive of a particular.


Suppose that we try to conceive of the Statue of Liberty. So I conceive of a tall greenish figure

on a large pedestal. Conceiving of a tall greenish figure on a large pedestal is unproblematic.16 But in

order for it to be a conceiving of the Statue of Liberty, we need to provide some way of ensuring that

what we conceive of is the Statue of Liberty and not some other tall greenish figure.





14Jonathan Dancy notes that in the 1710 edition, Berkeley continues: In truth the object and
sensation are the same thing, and cannot therefore be abstracted from each other.

5"Objects like numbers and sets, if they exist, will differ from ordinary contingent particulars in
this respect.

161 am here overlooking possible vagueness issues. If terms like 'pedestal' are vague it may be
that it is indeterminate whether there are any pedestals.










unicorns. So, were we to come across two creatures resembling horses with horns, each one with vastly

different DNA and internal organs, we would have no way of deciding which was the unicorn and which

was not. And, the reason we would not be able to tell which is in fact the unicorn is that neither is. We

should consider two claims that might be made about the conceivability of unicorns:


(r) I have conceived of unicorns.


(s) I have conceived that unicorns exist/ be existent.


The preceding digression is important to the discussion because it is not merely that the

existence of unicorns is impossible that should differentiate (r) from (s) but also why the existence of

unicorns is (at least according to Kripke) impossible that should help us distinguish the two.


The reason that unicorns are allegedly impossible creatures is that there is no description of

unicorns that would single them out as members of some species. And, since they do not exist, we

cannot fix the term 'unicorn' to any particular creatures and those things relevantly like them. So, any

attempt to conceive that unicorns exist would have to fail because there would be nothing to distinguish

our conceiving that unicorns exist from our conceiving that something superficially indiscernible from

unicorns exists.


But then can we conceive of unicorns? I believe that we can. Just think about Kripke's

argument that the existence of unicorns is impossible. In order to argue against their existence he must

be able to perform some mental act that at least resembles conceiving. There must be something

towards which his thoughts are directed. And that object is, at least in some sense, the fictional species

unicorn. In what way one is able to conceive of unicorns will be addressed later in the chapter. For

now, it will be enough that we can in some way conceive of impossible creatures so long as the content










nothing could exist to which the concept round square would apply and still succeed in conceiving of

round squares. But once one is aware that round squares cannot exist, he cannot truthfully and

sincerely utter something like (CTRS).


We should next consider an utterances like (CPE) and (CCRS):


(CPE) I have conceived that pigs exist (pigs to exist).


(CCRS) I cannot conceive that round squares exist (round squares to exist).


(CPE) is an assertion that one has performed some positive act which includes pigs existing. (CCRS)

reports either the failure to perform some positive act including round squares or, more likely, a

detection of some incoherence in the concept round square. At the very least, conceiving-that requires

that one not detect an incoherence either in the concepts the act of conceiving is about or in the

relations that the proposition designated by the that-clause puts those concepts in. Chalmers' prima

facie conceiving (discussed in section 2.2) is a type of conceiving-that which may be about impossible

objects. Even prima facie conceiving-that cannot be about objects that fall under concepts such that

grasping them puts one in the position to know they are incoherent. But one can conceive of concepts

and relations of concepts even if he knows them to be incoherent at the outset. Conceiving-of is an act

we perform when we consider hypothetical situations per impossible.


There is undoubtedly more to conceiving-that than can be explored in this study, and much of

that has been explored elsewhere.24 So I will at this point lay out the differences between conceiving-of

and conceiving-that and what the relation is between them. Conceiving-of amounts to first entertaining

a concept then drawing conclusions about at least some of the application conditions. Conceiving that p



24Much of this exploration has been published in the anthology Conceivability and Possibility
(eds. Gendler and Hawthorne).










(P112) may be stated in ordinary English. What (P112) says is that for any two distinct objects there is at

least one property that one has that the other does not.


The counterexample presented by B is a world empty except for two brass spheres which have

all their intrinsic properties in common. The two spheres are the same size, same density, same mass,

etc. And since they are the only two things in the world, they have all the same relational properties as

well. So although there are two objects they do not differ in any properties, which is exactly what (P112)

says cannot be.


One way to object to Black's counterexample is to propose that space is absolute. So, although

there are no purely qualitative properties that the two spheres differ in, they do differ in location.


If space is absolute, then Black's counterexample fails but it should not bear on what I am trying

to show here. I am not arguing that there are no real differences between distinct particulars. I am

arguing that there is nothing in our conceivings of particulars that would suffice to determine which

particular the conceivings are about. Even if we did include in our conceiving of the two spheres their

locations in absolute space, it would not be necessary that those two spheres be in the locations that

they are; one thing we do seem to know about ordinary particulars is that they could have been in

different spatio-temporal locations than they are in in fact.


Even though there may be some property that one sphere has that the other does not e.g. being

at such-and-such spatio-temporal location, it is not a necessary property. We may be able to think

about particulars in such a way as to pick them out uniquely. But when we conceive we need to leave

open that those properties instantiated by those particulars might have failed to be instantiated by

those particulars. And we further need to leave open that something else could have had the properties

of whatever particulars we are considering.










being which is omnipotent and omniscient.18 Anything which meets those criteria is God. Suppose one

were to say:


(G) I met God but he's not omniscient or omnipotent.


A natural response to an utterance of (G) would be that whoever or whatever the utterer met was not

God. That is because there are necessary and sufficient conditions on being God, as there are with

numbers. But ordinary particulars have accidental properties. And even if one believes that some

ordinary particulars have essential properties, (e.g. such as Humphrey having the property of being

human necessarily), it is still controversial whether there are properties are properties the having of

which suffices for an object to be the particular object it is.


There have been attempts at finding properties that individuate (suffice for being) contingent

ordinary particulars, but it is unclear that any of these attempts have been successful. And, even if one

had an account of ordinary particulars which included a non-controversial sufficient property, it is far

from obvious that when we think of ordinary particulars, we think about those types of properties. For

example, there have been theories about particulars which have as necessary and sufficient conditions

for being some object, 0, that O be composed of a certain group of sub-atomic particles. Even if such an

account is correct it does not seem like that is the type of thing we think about when we think about

ordinary particulars. And, even if we did think about ordinary particulars in that way, we do not have

the information regarding which elementary particles compose what in order for us to have in mind the

relevant properties necessary and sufficient for being any specific complex particular. That is, it may be

that one thinks in general that what individuates particulars is something about the particles of which



"There are arguably other descriptions which apply to God, like a being which is
omnibenevolent and the being who created the world. But whatever specific descriptions
one uses to pick out God is of no matter.










What is needed to show that there is a different act reported with 'conceive of' is a

demonstration that there are words or phrases that come after the complementizer 'of' that cannot be

paraphrased into a 'conceive that' construction. The content of (g) is also expressible as (h). But if we

look at (g), the of-clause attributes a property to a particular. But the of-clause need not be of that

form. If we look at (i),(j), and (k) below, we find that replacing 'of' with 'that' is problematic:


(i) I have conceived of George.


(j) I have conceived of ministers.


(k) I have conceived of being a minister.


Clearly, "I have conceived that George"/ "I have conceived that ministers"/ "I have conceived that being

a minister" are ungrammatical. One might suggest that (i), (j), and (k) implicitly include that the

referent(s) of the ordinary nouns and noun phrases after 'of' exist. So perhaps the content of (i), (j), and

(k) may also be expressed by (i'), (j'), and (k'):


(i') I have conceived that George exists.


(j') I have conceived that ministers exist.


(k') I have conceived that being a minister exists.


Looking back at (e), it is apparent that when the phrase after 'that' is in the present tense

'conceive that' statements like the three above are about what is epistemically open. But (i), (j), and (k)

do not necessarily convey that the existence of anything is epistemically open, even if the utterer of (i),


8There is a reading of (k) which reads the same as "I have conceived of my being a minister"
which would be expressible also with the 'conceive that' locution. I am here assuming that
there is a literal reading of (k) that is not elliptical for "I have conceived of my being a
minister". It is that reading which is important for the example.










we cannot perform the act of conceiving without analysis. My claim was, less ambitiously, that we need

to have a fuller understanding of conceiving in order to be sure that our arguments employing

conceiving do not fail to be formally valid or beg the question. So the first worry is not genuinely a

problem for this study.


The second worry stems from what is known as the 'paradox of analysis'. We cannot perform

any conceptual analysis without already grasping the concepts) in question. So, analysis cannot bring

us any understanding we did not already possess. But this is only a problem if we perform analysis with

the aim of acquiring new concepts or gaining some sort of new understanding. I do not purport to be

doing any such thing. Analysis, while not providing us with any new understanding, reveals what state it

is we are in with respect to some concept. One of my aims in chapter three will be to argue that there is

a species of conceiving that, while we understand what it is, has not been distinguished from other

types of conceiving or has been thought of as something else entirely. Consequently, this study should

not be affected by the paradox of analysis.


The third worry is that the method of conceptual analysis might fail in the instance of conceive

because we cannot make sense of iterated conceivings. If we were to conceptually analyze book we

might conceive of some book-candidate and consider whether the concept book applies to the book-

candidate. If we use the same method when we analyze conceiving, we would conceive of some

conceiving-candidate then consider whether conceiving applies to it. In order to make sense of this, we

must be the conceiver in this iterated act. But whatever we might learn from an iterated act of

conceiving can be learned just as well from a first-order act. For we can learn just as much through

investigation of what we conceive or fail to conceive as we do through considering what hypothetical

objects qualify as books.










of' construction as one can say with the 'conceive that' construction (and vice versa). But if it can be

shown that there are some things that can be said with reports using one construction and not the

other, then the two constructions are not used in all cases to report the same acts.


Before I get to the argument for the distinction between the two types of conceiving, it is

important that we first settle on a general form of report that does not prejudice our view of conceiving.

If we first look at the present tense uses of 'conceive that' as in (a), it is obvious that there is something

unnatural about them:


(a) I conceive that George is a minister.


(a) reads as if it were a performative like (b):


(b) I hereby state that George is a minister.


Whatever conceiving is, it is not a speech act, so it cannot be that (a) is a performative, at least not in

the normal way.


Perhaps (a) is a report of some pseudo-auditory experience of "I conceive that George is a

minister" the having of which is necessary and sufficient for the truth of (a). But of course one could

have a pseudo-auditory experience of "I conceive that George is a minister" without knowing what

'minister' means. So it cannot be sufficient for the truth of (a) that one have a pseudo-auditory

experience of (a). And while there may be something like a pseudo-auditory experience one normally

experiences before an utterance of (a), one need not have any, and even if one did it does not seem to

have any bearing on the truth or falsity of (a).


Since there is this tendency is to read (a) in a way that cannot be the correct way to understand

(a), we should use some other form of the 'conceive that' (and 'conceive of') construction for this

discussion.










2.3 Conclusion


In this chapter, I discussed the motivations for this study as well as some prima facie worries

that one might have about it. Philosophers have frequently made arguments that rely on conceivability

claims. But if we are unclear about what conceiving is or about distinctions between types of

conceiving, various problems can arise. Further clarification about conceiving would help us to avoid

those pitfalls. But there may be pitfalls in attempting to clarifying what conceiving is. So before the

discussion begins it is important to address the worries that might arise at the outset.


In section 2.1, I considered two problems that might arise with arguments relying on the

application of an unanalyzed concept. First, we might think that there is only one concept we are

applying, when in fact we are applying two closely related concepts by way of the same term. If that is

the case then arguments relying on conceivability claims may be invalid if different concepts are used in

each claim. Second, if conceivability is also expressed by some other term in the argument that looks as

if it expresses something distinct, then some arguments may be in danger of begging the question. For

example, Chalmers argues that conceivability entails possibility (and from there argues that

philosophical zombies are possible), but if conceivability


is nothing over-and-above possibility, then his argument begs the question as the conclusion is not only

contained in, but the same in content as one of the premises.


In section 2.2, I discussed three prima facie worries for this study. The first and third worries

stem from the methodology of conceptual analysis when applied to conceive, conceivable, etc. The

second worry is a worry about conceptual analysis in general.


The first worry is that if one does not fully understand what conceiving is, he cannot perform the

acts of conceiving necessary for analyzing what conceiving is. However, it was never my contention that

23










CHAPTER 5


CONCLUSION


In this study, I have examined the act of conceiving. The main conclusion of this study is that

'conceive of' denotes a substantively different cognitive act from 'conceive that'. This distinction is

useful in explaining confusions one might fall into regarding what is conceivable.


Before the main arguments for the distinction I motivated the study by suggesting that without

a better understanding of this act, arguments having premises about what we can conceive may prove

to be incorrect. There were two main reasons I suggested they may be incorrect. First, if 'conceive' is

used to report distinct acts from premise to premise, then arguments relying on conceivability claims

may be invalid. Second, if conceivability is the same concept as some other concept we thought to be

distinct, then some arguments relying on conceivability claims may be question-begging.


I later argued that acts reported by 'conceive of' and 'conceive that' are distinct acts. A failure to

distinguish the two might result not only in problems with validity and question-begging but further

confusions as well.


One confusion one might fall into is believing that it is conceivable that impossible objects exist.

I argued that the confusion lies in thinking one has conceived there to be some impossible object.

However, what one has in fact done in such cases is conceive of some type of impossible object.


Another way one might be confused regards the level of generality about which one may

conceive. I argued that conceiving cannot be done at the level of particulars. All conceiving is


of general types of things. I argued that a confusion about the level of generality about which we can

conceive might lead one to mistakenly argue for phenomenalism.










of those conceivings does not entail the existence of the creatures in question, factually or

counterfactually.


Of course, there is no reason to think that we are able to conceive of impossible creatures or

characters and not impossible objects. If we can conceive of unicorns we should also be able


to conceive of round squares. We cannot conceive that round squares exist or conceive of the existence

of round squares.


But why could we not conceive of round squares simpliciter? It certainly looks like we can in

some sense. Again, look at the linguistic evidence. Although we do hold that round squares are

impossible we also hold that if they did exist they would be round and square. So, they must in some

way be the objects of our thoughts about them. Undoubtedly, the concept round square exists.

Conceiving of round squares requires only that one gain an understanding as to what would be required

for some object to, per impossible, fall under both round and square.


In this section I have attempted to show that the acts reported by certain 'conceive of'

sentences differ substantively from the acts reported by 'conceive that' sentences. First, I demonstrated

that 'conceive of' and 'conceive that' are not intersubstitutable. Paraphrasing that-clauses in present

declarative tense and mood into present subjunctive will allow 'conceive that' assertions to report many

of the same particular acts that are reported by 'conceive of' assertions. But no amount of paraphrasing

will make 'conceive that' assertions report what is reported with a 'conceive of' assertion when the

direct object of the complement is a noun or noun phrase, with the paradigm examples being those in

which the noun or noun phrase denotes an impossible object.










one conceives that there be zombies, then it is possible for there to be zombies. And from this

possibility, one may conclude that physicalism is false. It is not necessary to go into detail about

zombies and physicalism2 to see that the argument will fail if when we think we have conceived that

there be zombies, we have failed to do so. The argument will also fail if when we think we have

conceived that there be zombies we have conceived zombies in some other way.


There is a general concern that if one's understanding of conceiving is impoverished in some

ways there are at least some arguments that rely on conceiving that one may not make. Suppose

someone, Luke, understands that currency is the physical representation of money, that American

dollars are the form of currency used in America and that Yen are the form of currency used in Japan.

Luke might understand what money is in some impoverished sense e.g., that it has physical

representations called 'currency', and that the type of currency used is in some way tied to political

entities. So, there might be some arguments that rely on grasping money that Luke could properly

make. But were we to know that Luke had this impoverished understanding of money, we would

probably not take seriously any arguments Luke makes about, for example, the global economy, nor

should we.


Similarly, in some instances we should not take seriously arguments relying on conceiving made

by people who have only an impoverished understanding of conceiving. So, it is important that we get

clear about conceiving. If our understanding of conceiving is incomplete, vague, or muddled then we

may not be able to make arguments relying on our ability to conceive. And, if our understanding is

incomplete, vague, or muddled it is unlikely that we would be able to tell when it is that we are justified

in making arguments that rely on conceiving.


2This will be done in section 4.3.










the two. In section 4.1, I consider the objection that all conceiving is conceiving-that. I argue that

conceiving-of and conceiving-that are substantively different and that there are common situations in

which we conceive-of and do not conceive-that. In section 4.2, I suggest that conceiving-of is a

precondition of conceiving-that. And, in so doing, I attempt to further explain the differences between

the two types of conceiving. In section 4.3, I suggest that a failure to recognize the distinction between

conceiving-of and conceiving-that may lead one to the false belief that he has conceived that zombies

exist. This, in turn, may lead one to argue that physicalism is false based because it is conceivable that

zombies exist (which is required for the zombie conceivability argument against physicalism) though in

fact one has only conceived of zombies.


I will close the study in chapter 5 with a brief conclusion in which I summarize the findings of the

chapters 2-4.
































2007 Kevin Savage










One might object that when we conceive we need not represent particulars with definite

descriptions in order to pick out those particulars. We need only stipulate that it is the Statue of Liberty

about which I am conceiving in order for it to be the Statue of Liberty about which I am thinking. But

one's conceiving of some particular cannot amount merely to stipulating that one has done so. For

simplicity's sake, we can consider a quasi-perceptual type of conceiving to illustrate that there are limits

to what one can stipulate. Suppose that one were to report that he had conceived of a specific shade of

blue, say robin's egg blue. We would not think that someone had successfully conceived of robin's egg

blue if he had in fact formed a quasi-perceptual image of a field of Titian (a shade of red), no matter

what one reports or stipulates.


Perhaps one might object to the example by arguing that quasi-perceptual conceiving requires

quasi-perceptual stipulation. One cannot stipulate merely linguistically. One must have an

understanding of the concepts involved in order to stipulate. And, where phenomenal


concepts are involved, one must be able to form a quasi-perceptual picture of a named color in order to

stipulate that one will be conceiving of that color.


But if that is what is required of stipulating that one has conceived of something, then the

stipulation is no different in type from the conceiving. So, it is of no use to argue that one can conceive

of a specific particular A merely because one stipulates that he conceives of A. One must further have

an understanding of A. The problem with conceiving of particulars is that it does not look like we

understand what ordinary objects like the Statue of Liberty are in addition to our understanding of the

many concepts under which they fall.


One might still think that we do not think about particulars as falling under some description.

But think about a supposed particular like God. When we think about who God is or would be, the way

we think about God requires that we think about God as falling under some description. God is that

50










LIST OF REFERENCES


Berkeley, George

1998 A Treatise Concerning the Principles of Human Knowledge edited by Jonathan Dancy (New York:
Oxford University Press).



Black, Max

1952 "The Identity of Indiscernibles" in Michael J. Loux (ed.), Metaphysics: Contemporary Readings
(New York: Routledge).


Chalmers, David

2002 "Does Conceivability Entail Possibility?" in Tamar Gendler and John Hawthorne (eds.),
Conceivability and Possibility (New York: Oxford University Press).



Hume, David

1978 A Treatise of Human Nature edited by by L. A. Selby-Bigge (New York: Oxford University Press).



Jackson, Frank

2002 "Finding the Mind in the Natural World" in David J. Chalmers (ed.), Philosophy of Mind: classical
and contemporary readings (New York: Oxford University Press).



Kripke, Saul

1980 Naming and Necessity (Cambridge, Mass.: Harvard University Press).



Perry, John

1979 "The Problem of the Essential Indexical," Nous Vol. 13 no. 1, pp. 3-21.


Yablo, Stephen

2000 "Textbook Kripkeanism and the Open Texture of Concepts," Pacific Philsophical Quarterly 81, pp.
98-122.










7. If God does not exist then God is an impossible being.


8. So, if God exists then God exists in every possible world and if God does not exist he exists in no

possible world.


9. Textbook Kripkeanism is false.


No problem would arise if one were to conceive of the existence of Pegasus and also of the non-

existence of Pegasus.12 No impossibilities arise from either the possibility of Pegasus' existence or the

possibility of Pegasus' nonexistence; Pegasus will exist in some possible worlds and not exist in others.

God is supposed to exist in all of them or none at all. It is impossible that he should exist in some but

not others.


Yablo's challenge is to the traditional view that conceivability entails possibility. So, his

argument is not directly aimed at a proposal like the one that I am working with currently. But, on his

way to arguing against the conceivability-possibility entailment, he does suppose that one can conceive

that there be impossible beings or situations. In order for his argument to work it must be either

impossible that God exists or impossible that he not exist. Consequently, in order for Yablo's argument

to succeed one must be able to conceive of some impossibility.


At first glance there are three options:


1. It is inconceivable that God exists.


2. It is inconceivable that God does not exist.


3. Both are conceivable.




12Or, at the very least, the same problem would not arise.














Table of Contents
ACKNO W LEDG M ENTS .......................................................................... ........................................ .... 4

L IST O F F IG U R E S ......................................................................................... ....................... ............ 6

A BST RA CT......................................................................................... ................... ............... ........ 7

1 INTRO D UCTIO N ............................................................ ............................................. ............. 9

2 MOTIVATION AND INITIAL WORRIES...................................................................... ...................... 13

2.0 Introd uctio n .................................................................................. ........................................... 13

2.1 Motivation................................ ........................................... 13

2.2 Initial W worries for the Project ............................................ ..................................................... 18

2 .3 C o n clu sio n ........................................................................................... .............................. ........... 2 3

3 THE REFUTATION OF THREE ASSUMPTIONS ABOUT CONCEIVING ..................................................... .. 26

3 .0 Intro d u ctio n .................................................................................. ............................................... 26

3.1 'Conceiving of and 'Conceiving that'................................................. ...................................... 28

3.2 Yablo and Conceiving of the Existence of the Impossible ................................ ...................... 39

3.3 Confusions That Could Lead to a Berkeleyan View...................................................... 47

3 .4 C on clu sion ......................................................................................... ................................ .........5 6

4 THE DISTINCTION BETWEEN CONCEIVING-OF AND CONCEIVING-THAT.......................... 58

4 .0 Introduction ........... ............................................................................. .............................................58

4.1 An Objection to the Distinction Between Conceiving-of and Conceiving-that....................... .. 59

4.2 The Relation Between Conceiving-of and Conceiving-that............................... ............. 62

4.3 The Implications of the Distinction for Zombies.................................... ............................. 67

4 .4 C on clu sion ......................................................................................... ................................ .........74

5 C O N C L U S IO N ............................................................................................................... ............ ....... 7 7

LIST O F R EFER EN C ES .............................................................................................................................. .... 79

BIO G RA PH ICA L SKETC H ............................................................................................... ........................... 80










obviously impossible object like the completely red and completely green ball. I do not think we can

even prima facie conceive that such an object exist, as even a minimal grasp of the concepts red and

green precludes a completely red and completely green ball existing. But, we can conceive of such an

object and draw conclusions about what features it would have to have were it possible for it to exist,

such as that it would be colored, have a surface, etc.


In this section, I have looked at what I suggest are some of the confusions people have when

they conceive. When people think that they have conceived that some impossible object exists at least

one of the explanations is that one has confused conceiving of the impossible object for conceiving that

the impossible object in question exists (or conceiving it to exist). What one can conceive as existing is

limited by what is possible. But what one can conceive of is not so limited.


I then considered an objection not to the suggestion that one can conceive of impossibilities but

to the suggestion that one cannot conceive that impossible objects exist. The objection is one suggested

by Stephen Yablo. Yablo suggests that we can both conceive that God exist and conceive that God not

exist. Since God, if he exists, is supposed to be necessary, either his existence or non-existence is

impossible. I attempted to show that this apparent problem can be explained away with the distinction

I suggest between 'conceiving of' and 'conceiving that'.


Lastly, I considered an objection that the kind of solution I suggest has already been offered in a

different way. Yablo's confusion could also be explained away with Chalmers' distinction between prima

facie and ideal variants of conceiving. And so, one might think that the distinction between conceiving

of and conceiving that is nothing over-and-above the distinction between prima facie and ideal

conceiving. I argued that conceiving of differs in type from prima facie conceiving in that one may

conceive of some types of obviously impossible objects which one cannot even prima facie conceive.










conception noun The action or faculty of conceiving in the mind, or of

forming an idea or notion of anything; apprehension,

imagination.


conceive verb to take into, form in, the mind


conceiving verb forming in the mind


If we look at the list above, 'conceivable' and 'conceiver' both explicitly have 'conceive' in their

definitions. So on their own, those definitions of 'conceivable' and 'conceiver' will not be helpful in

understanding the definition of 'conceive' because we would want a definition to provide us with an

explication of 'conceive' that does not have 'conceive' in the definition, for that would be an obviously

circular definition. If we look at the definition of 'conceive' it is not explicitly circular.


While the circularity resulting from the definition of 'conceive' is not precisely the same as the

one resulting from particular definitions of 'conceivable' and 'conceiver', it is problematic. What we

want to understand in trying to understand what the act of conceiving is, is not something only

linguistic. My contention is only that there is a problem with understanding the term 'conceive' but that

there is a problem with understanding what it is to take something into the mind in the relevant way

and the ways that we conceive or take something into the mind.6


If 'to take into...the mind' is strictly synonymous with 'conceive', then understanding either term

or phrase would require one to grasp precisely the same conceptss. So, in order to grasp conceive one

would have to first grasp conceive. This appears to be viciously circular. If in order to understand A one


6Additionally, 'to take into the mind' may be a phrase that describes some of what it is to
conceive, but which does not provide necessary or sufficient conditions on conceiving and
which is vague. One might describe perceiving as a kind of 'taking into the mind' but it is
wrong to think that perceiving is conceiving (although Berkeley did not think so).










they are composed, but it is unlikely that there are any particulars for which one can enumerate which

particles that particular is composed of.


So, while some conceivings may be about certain odd particulars like God or numbers, it is

because what they are can be captured by a definite description which we do in fact have in mind when

we think about them. And those definite descriptions which apply to these odd particulars pick out

these particulars uniquely and apply to those particulars necessarily. With ordinary particulars, it is not

obvious that there are any descriptions which uniquely pick out those particulars and apply to those

particulars necessarily. Even if there are it is unlikely that we have such descriptions in mind when we

think about them.


To demonstrate the point that there is no necessary description under which we can uniquely

pick out particulars I will look at Max Black's argument against the Principle of the Identity of

Indiscernibles. Black's argument takes the form of a dialogue between two characters, A and B. B

presents a counterexample to the Principle of the Identity of Indiscernibles (Pll):


(PII) For all objects x and y and properties F, if (x has F iff y has F) then x = y.


(PII) provides a condition sufficing for the identity of particulars. The principle is difficult to state in

ordinary English as in (P1l)19 but it may be better put in ordinary English if we first put it in the logical

equivalent of (PII), (P112):


(P112) For all objects x and y, if x # y then ~for all properties F(x has F iff y has F)






191t is natural to say that for any two objects which have all the same properties, they are
identical. But then there is a problem because there is only one object, not two as first
stated.










5'. I can conceive of a minimal physical duplicate of our world which is not a duplicate

simpliciter of our world.


Premise 6 would remain unchanged:


6. If it is conceivable that there be a minimal physical duplicate of our world which is not

a duplicate simpliciter of our world then it is possible that there be a minimal physical

duplicate of our world which is not a duplicate simpliciter of our world.


Whereas the statement in premise 5 was the antecedent of the conditional of premise 6, 5' is not the

antecedent of 6. So, then it does not follow that zombie-worlds are possible. Likewise, it does not then

follow that physicalism is false.


The question then is: what is the substantive difference between 5 and 5'? It is the difference

between conceiving-of and conceiving-that has been the main focus of this study. Conceiving-of is not

evaluative in the way that conceiving-that is. Conceiving-of consists of entertaining a concept (coherent

or not) and drawing conclusions as to what falls under the concept, as we do with conceptual analysis.

Conceiving-that requires that one not detect that the concepts are incoherent, while conceiving-of does

not. Because conceiving-of does not require this failure to detect incoherence, one can conceive of

impossible objects. Further, since propositions are not the objects of conceiving-of nothing one

conceives of can be either true or false, so conceiving-of cannot have metaphysical implications. So,

whether or not zombies are possible, they are conceivable in the sense that they may be conceived of.


To illustrate, one may conceive of fictions. And even if one does not hold that fictions are

impossible in virtue of being fictions, there are fictions which are impossible and yet we still do not fail

to conceive in the way that we conceive other fictions. One might write a story in which all of the

physical laws are explicitly stated. Also in this story is a naturally blond-haired woman named Callie










3.2 Yablo and Conceiving of the Existence of the Impossible


Can we conceive the existence of only what is possible? A complete answer to this question

would be beyond the scope of this paper. Regardless, most accept that there is at least some

connection between conceivability and possibility. Conceiving seems to give us some epistemic access

to possibility.'0


However, it seems clear is that there are at least some things we do not think can be conceived

because they are conceptually impossible. The types of things we exclude are things like the existence

of impossible objects and situations which we know are logically impossible. We do not think that

anyone can conceive of the existence of round squares or dogs that are not mammals. And we think it's

inconceivable for a stone to be bigger than itself.


But what about people who do think they can conceive of the existence of round squares and

stones larger than themselves? That is, how is it that we understand what is going on with those

people? We have four options:


1. They are genuinely conceiving of some impossibility and we are mistaken that such things

cannot be conceived of.


2. Those things we thought were impossible are not impossible.


3. They are not conceiving anything at all.


4. They are conceiving of something but are mistaken about how to describe what it is they are

conceiving.



10Whether or not conceiving is a reliable guide to possibility is another matter.










replaced by a noun or noun phrase. I considered the suggestion that 'I have conceived of x' is elliptical

for 'I have conceived that x exists (x to exist)' and found that although the former may seem to be

elliptical for the latter for some substitutions for 'x', the former is obviously not elliptical for the latter

for all substitutions for 'x.' The substitutions for 'x' which most obviously do not admit of the ellipsis are

those substitutions which denote impossible objects. And from that, it should follow that 'conceive of'

sentences are in general not elliptical for 'conceive that' sentences.


One could argue that the problem lies in thinking that we can conceive of impossible objects.

Hume made this point in A Treatise of Human Nature:


'Tis an established 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. (Book 1. Part 2.Section 2. Paragraph 8)



Hume argues, as many have, that what is impossible is inconceivable. From the passage above,it is clear

that Hume's view is that we cannot conceive of conceptual impossibilities.22 The conceptual

incoherence ofx leads to an inability to conceive ofx. Whether or not we can form an idea of a golden

mountain depends on whether the concept golden mountain is coherent. But merely detecting the

coherence of golden mountain is not enough to conceive of a golden mountain on this view. Conceiving

of a golden mountain is something we can do in virtue of golden mountain being a coherent concept.





22Just what types of possibility and impossibility there are is beyond the scope of this paper. For
purposes of this study, I am assuming that there are conceptual, logical, and nomological
possibilities. Whether conceptual and logical possibility are the same or not is another issue
beyond the scope of this paper. However, nothing important should hang on whether or not
conceptual and logical possibility are distinct notions.



















CHAPTER 2


MOTIVATION AND INITIAL WORRIES


2.0 Introduction


In this chapter, I will discuss preliminary considerations for this study, laying out the motivations

and addressing some possible problems at the outset. In section 2.1, I discuss more fully the motivation

for this study and various problems that might arise without a more complete understanding of the act

(or acts) of conceiving. Those problems are that a less-than-full understanding of conceiving may lead

us to make arguments that are either not formally valid or are question-begging. In section 2.2, I

explore the prima facie problems that might be posed for this study. There are two prima facie

problems relating to apparent circularity: the paradox of analysis and what might be called "the paradox

of performance." There is also a third possible problem which is that one might have to perform an

iterated act of conceiving in order to analyze conceive.1 I conclude that none of these prima facie

problems, whether serious or not, should adversely affect this study.


2.1 Motivation



Let me being with looking at a contemporary example of an argument that relies importantly on

conceiving to arrive at its conclusion. In "Does Conceivability Entail Possibility?" (Chalmers, 2002), David

Chalmers has argued that (the right kind of) conceivability entails possibility. So, if Chalmers is right, if

1Words in italics will be used to denote concepts as in conceive, conceiving, etc.










Smith fail to apply top Smith. There is nothing in the content of our acts of conceiving that would

uniquely pick out any particular in all possible circumstances. Particulars exist. I am not promoting an

eliminativist thesis. Rather, I hold that the contents of our conceivings represent the world in such a

way that they cannot be about particulars as such.


In section 4.1, I will consider an important objection to the view that conceiving-of is a

substantively different type of conceiving from conceiving-that. Specifically, one might argue that

conceiving-of, in the instances it differs from conceiving-that, is not a type of conceiving at all. I argue

that we do recognize this difference in normal parlance and in logic as well, at least implicitly. I conclude

that the suggestion that substantively different conceivings-of are not conceivings at all is mistaken.


In section 4.2, I will attempt to show that conceiving-of is a precondition on conceiving-that and

further discuss what is required for conceiving-that. What is required to conceive-that is more involved

than what is minimally required for conceiving-of, but one may not conceive-that without first

conceiving-of.


In section 4.3, I will suggest that the failure to recognize this distinction could be behind the

apparent conceivability of philosophical zombies. A full argument that such a confusion is going on in

zombie arguments is beyond the scope of this thesis. However, I will demonstrate how such a confusion

could lead one to believe that philosophical zombies are conceivable.


4.1 An Objection to the Distinction Between Conceiving-of and Conceiving-that



One might argue that there really is no reason to think that in addition to the type of conceiving

normally discussed (conceiving-that) that there is also some other type of conceiving (conceiving-of). I

argued in chapter two that there is a substantive difference between conceiving-of and conceiving-that

which is demonstrated by the fact that we do make reports of the form 'I have conceived of x' where x is

59










ACKNOWLEDGMENTS


I would like to thank Sara, without whose support I could never have stayed motivated. Thanks

also go to Kirk Ludwig, whose advice and ability to reason were invaluable. I also thank Gene Witmer,

without whom I might not be in philosophy, for his admirable ability and integrity. And lastly, thanks go

to Michael Jubien, who very graciously stepped into my committee at the 11th hour.










at a higher level of generality then there is no conceptual impossibility as there is with some specific

particular being both conceived and unconceived.





3.4 Conclusion



In this chapter I have looked at some mistakes that can be made when one allegedly conceives.

Each mistake I looked at led to some conclusion about the range of things which can be conceived.


In section 3.1, I suggested that a mistake that has been made, particularly in recent discussions,

is in considering only conceiving about propositional contents. I argued that there are mental acts

reported by sentences using the 'conceive of' locution that are substantively different from any acts that

can be reported by sentences using the 'conceive that' locution. Those acts that differ are the ones

reported by 'conceive of' followed by an ordinary noun or noun phrase. I considered the suggestion that

(i)'conceive of A', where 'A' is an ordinary noun or noun phrase, is elliptical for (ii)'conceive that A exist'.

I concluded that the (i) is not elliptical for (ii) because we can in some way conceive of impossible

objects (and believe that those objects are impossible). But the way we can conceive of impossible

objects must be a way which does not require us to conceive that they exist.


In section 3.2, I considered an objection that the existence of impossible objects is conceivable.

The objection came from an argument made by Stephen Yablo against Textbook Kripkeanism. Yablo's

argument presumes that both the existence and nonexistence of God are conceivable. One of either the

existence or nonexistence of God is necessary and one of them is impossible. So, if both of them are

conceivable, something impossible is conceivable. I argued that one might be misled into believing both

the existence and nonexistence of God are conceivable if he is confused between conceiving of and

conceiving that. Lastly, I argued that in spite of appearances, the distinction between conceiving of and

56










absurdum is, of course, not a perfect model. Reductio ad absurdum works with propositions but

conceiving-of manipulates concepts rather than propositions. But if we can make sense of an intuitive

notion of conceptual entailment, we see that the two are not far off from each other.


Of course, one could object that while we do perform this type of act, it is not a form of

conceiving. Instead, it should be called 'considering' or 'supposing'. My reply is that it is fine to call the

act whatever one wants to call it. However, as I will suggest in section 4.2, there is an intimate

connection between conceiving-of and conceiving-that which warrants putting both under the heading

of conceiving. And, as I have suggest throughout this study, we do perform both of these acts and use

forms of the word 'conceive' to refer to both acts. In section 4.3, I will suggest that the fact that we do

use 'conceive' to refer to both acts may have led to a confusion in contemporary philosophy of mind in a

matter of some importance.


In this section, I have considered the objection that there really is no distinction between

conceiving-of and conceiving-that. I have argued that we do perform something which looks a lot like

conceiving-of with the reduction ad absurdum. I also argued that we do conceive of impossible objects.

The evidence that we do so was that we do think that the contents of conceiving reports of different

types of impossible objects are different reports.


4.2 The Relation Between Conceiving-of and Conceiving-that


I have described conceiving-of thus far as an act that puts us in the position to understand the

application conditions for some concept, whether it is simple or complex. Because conceiving-of does

not require that one conclude that the concept is coherent, one may conceive of types of impossible

objects. But one cannot conceive that impossible objects exist.












CHAPTER 3

THE REFUTATION OF THREE ASSUMPTIONS ABOUT CONCEIVING

3.0 Introduction




The goal of this chapter is to establish that there are two importantly different types of

conceiving which are reported using the locutions 'conceiving that' and 'conceiving of' respectively, the

latter of which is not usually distinguished from the former, and to show that conceiving of in the

relevant sense is never of particulars but only types of things, and that the types of things we can

conceive involve inter alia impossible objects. I show in developing this account how this distinction can

be used to diagnose some problematic objections to it and some other well-known philosophical

arguments.


The method of this chapter is to examine the sorts of mistakes we make when we conceive or

attempt to conceive as a way of articulating what is involved in it. I approach this by examining the

types of mistakes people make when they report conceiving something. I consider whether the

mistakes people make in these reports are about the content of the act of conceiving or about the act in

which they were engaged. Looking at mistakes in conceiving in this way will illuminate both the range of

what is conceivable, what conceiving is more generally, and whether there are types of conceiving

heretofore unconsidered.


The plan of the chapter is as follows.


In section 3.1, I look at the 'conceive that' construction and consider whether it is the

construction we should be using exclusively. I argue it is not and in particular that the 'conceive of'

construction expresses a substantively different mental act from the mental act one reports with the










conceivable and that we can conceive of particulars (e.g. God, people we know, etc.). In chapter 3, I will

argue that all of these assumptions are false.


First, I will argue that reports using 'conceive of' and 'conceive that' are reports of substantively

different cognitive operations. I will perform a series of comparisons between 'conceive of' and

'conceive that' sentences in order to show first that 'conceive of' and 'conceive that' are not

intersubstitutable and thus not synonymous. I will use further comparisons between sentences of both

sorts to demonstrate that there are some 'conceive of' sentences that cannot be paraphrased as

'conceive that' sentences. This gives us strong reason to think that 'conceive of' and 'conceive that'

report different types of operations. The argument that 'conceive of' and 'conceive that' report

different operations is intertwined with my denial of the second assumption: that we cannot conceive

what is impossible.


Second, I will argue that we can conceive of impossible objects, but that we cannot perform any

cognitive acts reported with 'conceive that' sentences which are about impossible objects. This is

because conceiving-that requires that we think of the objects we are conceiving as existent in some

sense. For example, we can conceive of round squares but we cannot conceive that round squares exist,

that there are houses adorned with round squares, or that the president was born in a round square

cupola. I will consider an objection not that my requirements for what we can conceive is too weak but

that they are too strong. The objection supposes that we can conceive that impossible objects exist or

conceive of them as existing. I will suggest that what may make it seem conceivable that impossible

objects exist is a failure to distinguish between conceiving-of and conceiving-that.


Third, I will argue that we cannot conceive of particulars or conceive that particulars exist. Of

course, particulars do exist. But the way we must think about particulars in acts of conceiving does not

uniquely pick out any particulars. Conceiving requires us to leave open that particulars could be or could










the same word are examples of the type of mistake that might be made. This is the type of mistake that

is highlighted by puns.4


Sometimes we make mistakes that are easily discoverable. Suppose that one were to overhear

a conversation in which one man, explaining to another why his wife could not make it to the party, says

"Mary is a little hoarse." A man overhearing the conversation comments to the man with the ill spouse

"I have an Appaloosa myself." Funny or not, these types of mistakes do get made even in situations in

which we can give somewhat clear analyses of the concepts in question. We can easily identify the

mistake in the example just given. For one, 'horse' and 'hoarse' are spelled differently, so anyone

reading an account of the scenario just described, even someone who does not know the difference in

meaning between the two, will immediately be predisposed to believe that they 'horse' and 'hoarse'

mean different things. And, anyone who does understand what each term means should not have any

trouble noticing the difference between 'horse' and 'hoarse' as the two are so different in meaning.

Even if one were to misspell one of the terms in an account above, the difference would still be easily

spotted because the contexts of utterance of each term would cue a reader to the mistake.


But differences between what terms mean need not be so obvious. Sometimes differences may

go unnoticed. Consider 'unintended' and 'accidental'. Although there is a difference between what the

two terms mean, both terms may be used in almost all of the same contexts and they are often used

interchangeably.


If instead of with two terms, what is meant by 'unintended' and 'accidental' were expressed

with a single term, it would be incredibly difficult to notice that there were two different concepts at

work. And if 'conceive' is a polysemous term which expresses multiple distinct, albeit inter-related


4While one does not mistakenly make a pun (in fact, one would not be making a pun if the act in
question were unintentional), one plays at making a mistake with a pun.










CHAPTER 4

THE DISTINCTION BETWEEN CONCEIVING-OF AND CONCEIVING-THAT


4.0 Introduction



There are three objectives of this chapter. First, I will further defend the distinction between

conceiving-of and conceiving-that. Second, I will discuss the relation between conceiving-of and

conceiving-that. And third, I will suggest how the distinction between these two types of conceiving

might help explain away the apparent conceivability of zombies existing.


In chapter 3, I argued for a number of different proposals regarding the range of those things

that can be conceived. When philosophers appeal to what can be conceived in arguments, the type of

conceiving they normally focus on is reported with the 'conceive that' construction (or similar

constructions with a sentential complement). I argued that there is another type of conceiving reported

with the 'conceive of' construction. Conceivings-of may seem to be about ordinary objects but not be

about the existence of those objects. So, one may be able to conceive of some type of thing without

conceiving that that type of thing exists. Because we may conceive of types of things without conceiving

that they exist, we can conceive of types of impossible objects.


I also argued that we cannot conceive anything about particulars qua particulars. Conceiving

requires that we leave open the possibility that things be different than they are or than we believe

them to be. If one conceives that, for example, Smith is taller than six feet tall, his success in conceiving

that Smith is taller than six feet tall does not depend on Smith's actual height. In order for it to be Smith

that we have conceived to be six feet tall, we must be able to differentiate between Smith and, for

example, Jones in our acts of conceiving about Smith and Jones. The way in which we pick out Smith is

by way of some descriptionss. But we can conceive that any or all descriptions which in fact apply to










2. Possible(a) Possible(a) (1)


3. Possible(a)


4. Possible(a) (2,3)


The argument is valid, but question-begging (and thus not informative), for it assumes the conclusion as

a premise.


In this section I have tried to motivate the project by explaining several ways in which one may

be confused about what conceiving is and why this is important. I have shown that if we are confused in

the ways described above, then arguments we take to be valid may not be. A theory of conceiving will

help guard against these types of confusions; what we do not know about conceiving can hurt us.


2.2 Initial Worries for the Project



In this section I will look at some prima facie problems for an examination of conceiving that

stem from the nature of conceiving and philosophical methodology. This study will in some instances

examine conceiving by means of thought experiments. These thought experiments will sometimes have

conceiving and conceivability as the objects) of the thought experiments. They will be used to tell us

something about the limits of conceiving and about what conceiving is.


There is at least a prima facie worry that might arise about this methodology. I will be

examining conceiving and related concepts by way of conceiving. So, there is at least a hint of circularity

looming in the wings for my proposed analysis.


In section 2.1, I argued that there could be problems with particular arguments which rely on

acts of conceiving without a fuller understanding of conceiving. I will have to perform the thought

experiments meant to provide a fuller understanding of conceiving without that fuller understanding I

18










used to motivate the project. So, I will in essence be committing the philosophical sin which the pursuit

of this project was to enable us to avoid. But it is not my contention that we have absolutely no starting

point. We can perform the act of conceiving without an analysis. But without the analysis we may not

always have the tools to determine when it is we have conceived in the way that is relevant to what our

aim is. I contend that we perform a type of conceiving that has not yet been discussed in the literature.

And, although we do perform this act, we have not yet examined it so we do not recognize it. If we

could not perform any specific mental acts without analyses, we would never have any acts subject to

analysis. So, this worry is really a red herring, but I think it important to get it on the table nonetheless.


The second worry relates in a general way to the first although more serious. Part of what will

be done in this study is both a conceptual and linguistic analysis. But analyses are subject to the

paradox of analysis. Below are the definitions of 'conceive' and its forms from the Oxford English

Dictionary:


Word Part of Speech Definition


conceivable adjective that can be conceived of or thought of; imaginable,

supposable


conceived adjective admitted into, or originated in, the mind; imagined, thought

of, etc.


conceivement noun conception


conceiver noun one who conceives


conceiving noun conception



SDefinitions are the first relevant definitions for each term in the OED.

























Figure 1. A Penrose Triangle










I then considered the objection that conceiving-of is not conceiving at all. I concluded that this

objection is mistaken. One might call conceiving-of by some other name but it has an intimate relation

to conceiving-that which warrants the use of the term 'conceiving'.


When we conceive-of we entertain a concept, whether simple or complex, and figure out

application conditions for that concept. This type of act is modeled, albeit inexactly, by reduction ad

absurdum. In a reduction we perform the operations necessary to draw a contradiction from an

assumption. Conceiving-of is analogous to the figuring out part of reduction ad absurdum but is

disanalogous in that conceiving-of does not aim at drawing a contradiction.


When we conceive that p we must know at least some of the application conditions of the

concepts in p. And the mechanism by which we come to know these application conditions is

conceiving-of. So, conceiving-of is a precondition on conceiving-that.


With this distinction, we might be able to diagnose argumentative mistakes we did not earlier

believe to be mistakes. Or, we might have a different understanding of the mistakes we do think have

been made in arguments. I suggested that this distinction might be useful in diagnosing the mistake

made when one apparently conceives that zombies exist. Whether or not zombies are conceivable, it is

worth exploring whether the mistake I suggest is made with zombie arguments, as with other

arguments that rely on conceivability claims.










just what we conceive but what we see, hear, taste, etc. We are even likely to be mistaken in some

instances about our beliefs.


We must then wonder what the nature of the mistake is. We are going to assume that the

person mistaken has at least a basic understanding of what conceiving is so that he does not fail to

conceive of anything. Further, we must assume that whoever makes the claim:


(RS) I am conceiving of a round square's existence.


is a competent speaker of the English language who understands all of the term in question. That is, he

does not report RS in virtue of not knowing the meaning of the term 'round' (or 'square' or 'existence',

etc.). If one sincerely asserts RS, is a competent speaker of the English language but fails to conceive of

the existence of a round square then he must have attempted to conceive of round square but failed to

do so:


(MC) For all speakers S and languages L, if S is a competent speaker of L, sincerely utters 'I am

conceiving of x', then S sincerely attempts to conceive of x, and if S fails to conceive of x, S


succeeds in conceiving of something else.


According to the above principle, if one were to report that he conceived of a round square and failed to

do so (which I am holding by hypothesis that he must) he would not of course be conceiving of a round

square, but he would succeed in conceiving of something.


A principle similar to the one above is suggested by Saul Kripke. Stephen Yablo calls a similar

view 'Textbook Kripkeanism':


A lot of people appear to have drawn the same 'good news-bad news' lesson from their reading
of Saul Kripke on conceivability. The bad news is that conceivability evidence, particularly of the
'conceptual' or 'a priori' sort, is highly fallible. Very often one finds a statement E conceivable,










Round square is not a coherent concept, yet I have argued that we can conceive of round

squares. What we cannot do is conceive that round squares exist (conceive round squares to exist).

Hume says we cannot clearly conceive of impossibilities. I would take this to be saying that we cannot

ideally conceive that there be impossible objects. And that is not in conflict with anything I have

suggested in this study. One might still object that even if that is what Hume has in mind, there is no

reason to think that there is this distinction between conceiving-of and conceiving-that.


But we do conceive of impossible objects at some times, even while admitting that they are

impossible. When we talk about round squares we have done something which allows us to sincerely

say about them what we do. Even though round square is an incoherent concept, we can make sense of

what it would take for something to be a round square. We can draw conclusions about what would be

true of round squares were they to exist. If we could not conceive of round squares then we should

think that (CORS) and (COET) have the same truth conditions:


(CORS) I have conceived of a round square.


(COET) I have conceived of an elliptical triangle.


But it is not obvious that (CORS) and (COET) have the same truth conditions. Further, I have the

intuition that the truth of (CORS) require an utterer to perform an act different from the act required for

one to truthfully utter (COET). I think we have this intuition in spite of the fact that we know at the

outset that both round squares and elliptical triangles are impossible.


In logic, we have a model of this type of conceiving with the reduction ad absurdum. When one

performs a reduction, one withholds judgments about the assumption of the reduction until it has been

revealed that the assumption implies a contradiction. It is in this way that we conceive of

impossibilities. We grasp the concept or concepts and from there draw conclusions. Reductio ad












BIOGRAPHICAL SKETCH

Kevin Savage received his B.A. in philosophy from the University of Florida in December 2003.

Kevin is currently a graduate student in the philosophy department at the University of Florida

specializing in metaphysics, philosophy of language, and philosophy of mind.










would also be creatures which are minimal physical duplicates of us. Some have argued that it is

conceivable for such creatures in such a world to lack mental states. These minimal physical duplicates

of us that lack mental states are called 'zombies'.


Zombies are also behavioral duplicates of us. They appear to walk. They seem to laugh and cry,

but are not amused by what we would think of as jokes and are not saddened by what we would call

'tragedies'. So, a zombie world would be behaviorally and physically just like ours.


Although he was not the first to hold that conceivability entails possibility (of some kind) David

Chalmers has argued that conceivability entails possibility (Chalmers, 2002). Chalmers is careful to

explain that it is only a specific type (or types) of conceivability that entails metaphysical possibility. But

all types of conceiving that Chalmers discusses are varieties of conceiving-that.


Jointly holding that zombies are conceivable (in the right way) and that (the right kind of)

conceivability entails possibility will commit one to the falsity of physicalism.26 Below, I will give a

simplified version of the argument:


1. Any world that is a minimal physical duplicate of our world is a duplicate simpliciter of our

world. (thesis of physicalism as in (III))


2. It is conceivable that there be a a minimal physical duplicate of our world in which there are

zombies.


3. It is conceivable that there be a minimal physical duplicate of our world which lacks mental



26Chalmers argues that one is actually committed to either the falsity of physicalism or the truth
of panprotopsychism. However, it is unclear whether panprotopsychism should be
considered a type of physicalism. Regardless, it is not the view that philosophers generally
have in mind when they talk about physicalism.










(e) I have conceived that George is a minister.


(e) is true or false solely in virtue of whether or not one has performed the mental act of conceiving that

George is a minister.


Now I turn to the difference between those sentences that take the complementizers 'that' and

'of' after 'have conceived'. First, I will look at whether 'have conceived of' and 'have conceived that' are

intersubstitutable salva veritate.


(e) I have conceived that George is a minister.


(f) I have conceived of George is a minister.


Obviously, 'conceived of' and 'conceived that' are not intersubstitutable, as evidenced by (e) and (f).

While (e) is truth valuable, (f) is not a grammatical sentence, and so not truth valuable. It follows that

the two phrases are not synonymous.


If there are things that can be said with sentences using the 'conceive of' construction that

cannot be said with any sentences using the 'conceive that' construction then there is further reason to

think that there is a substantive difference in the acts reported by each construction.


A mere failure of intersubstitution will not show that there are two substantively different

mental acts reported with each locution. So I will attempt to demonstrate that the way in which

'conceive of' and 'conceive that' fail to be intersubstitutable cannot be bridged by changing other terms

in the sentences.


First, we should look at whether what is said with (e) can be said with any sentence using

'conceive of'. As seen with (f), 'conceive of' cannot be substituted for 'conceive that' in (e). But there










facie conceived there to be a greatest prime number or merely conceived of a greatest prime number.

Whichever is the case, it is important to notice that grasping greatest prime number does not by itself

put one in the position to detect the incoherence.


In this section, I have looked at what implications the conceiving-of/conceiving-that distinction

might have for the zombie-argument against physicalism. I argued that there could be a confusion

between conceiving-of and conceiving-that in the argument. If there is such a confusion then it would

be particularly hard to detect as one may merely conceive of zombies and possess all of the non-

inferential knowledge one would need to conceive that zombies exist. But like some fictions and

impossible drawings, we can conceive of zombies even if they are impossible.




4.4 Conclusion



In this chapter, I have discussed the significance of the distinction between conceiving-of and

conceiving-that. In section 4.1, I considered the objection that the distinction does not really exist. I

argued that we do perform an act that we call 'conceiving' that allows us to conceive impossible objects.

And conceiving-of is that act which allows us to conceive impossible objects.


In section 4.2, I discussed the relation between conceiving-of and conceiving-that. Conceiving-of

is a precondition of conceiving-that. I also further discussed the requirements on conceiving-that. In

order to conceive that p27 one must: (1) first conceive of all x's that p is about.; and (2) not detect any




27Again, we should keep in mind that one may prima facie conceive that p even if p is
contradictory. Hence the requirement is that we not detect an incoherence rather than
there being no incoherence or no detectable coherence, regardless of the amount and
quality of rational reflection.










Perhaps we can provide some definite description of the Statue of Liberty. But for any definite

description we give of the Statue of Liberty, that description might have failed to apply to the Statue of

Liberty. It might have been shorter, constructed somewhere other than Ellis Island, designed earlier or

later than it was, etc. It may be that there is no problem with picking out a particular like the Statue of

Liberty with some mental acts, but I want to suggest that conceiving is importantly different. When we

conceive we are directing ourselves toward modal concerns. So we must leave open that particulars

could have been different than they are.


We might think that the Statue of Liberty falls under the definite description "the actual tall

greenish statue on Ellis Island." Of course, it cannot fail to be the case that the Statue of Liberty is the

actual tall greenish statue on Ellis Island. But when we conceive that the Statue of Liberty be

somewhere other than Ellis Island we are considering whether we think that situation could have been

actual. But if we represented the Statue of Liberty as "the actual tall greenish statue on Ellis Island" we

could not have conceived that the Statue of Liberty actually be assembled anywhere other than Ellis

Island. In possible worlds speak, if our modal


considerations are about anything, they should be about how the actual world could be, not about how

some other world is.


Of course, objects like numbers (if they exist) would be exceptions, as it is uncontroversial that

they fall under definite descriptions that could not have failed to apply to them. Another notable

exception is in first-person conceivings. Whenever I conceive of myself first-personally, it is undoubtedly

me that my conceiving is about. And that is because I do not need to represent myself with some

definite description in order to guarantee that it is me about whom I am conceiving.17




17Perry, John. "The Problem of the Essential Indexical", Nous Vol. 13 No. 1 (1979): pp. 3-21










To more clearly see the danger, we may examine the structure of the argument that

philosophical zombies are metaphysically possible schematically3:


1. (Vx) (Conceivable(x) Possible(x))


2. Conceivable(a) Possible(a) (1)


3. Conceivable(a)


4. Possible(a) (2,3)


The argument above is obviously formally valid, assuming univocality of terms. But of course an

argument's conclusion may be false even if the argument is formally valid, if one or another premise is

not true.


The charge against arguments like Chalmers' that I wish to examine is not that the third premise

is false. Normally, one would argue against Chalmers' argument by arguing that the existence of

zombies is inconceivable. The worries I want to suggest are a bit different and not aimed only at

Chalmers' argument but any argument relying on acts of conceiving. Any argument which relies on

some fact(s) about what we can conceive will be adversely affected by an incomplete understanding of

the act of conceiving.


The first worry is that one might be using different concepts from one step of the argument to

the next. Simple misunderstandings stemming from the use of homophones or alternate definitions of





31n these examples, 'conceivable' and 'possible' will be treated as predicates, for simplicity's
sake. And for present purposes, 'x' and 'a' can take anything whatsoever for their values.
These examples are only meant to demonstrate problems that can arise from arguments
superficially of this form.










If there is any limit to what is conceivable, it is most likely the existence of impossible objects.

So, if option 1 is correct then there is likely no meaningful string of words that does not correspond to

something that can be conceived. In itself, this is not an objection but an observation that should make

one uneasy about going with option 1. "It is conceivable that P" would be nearly synonymous with "'P'

is a meaningful sentence." Conceiving is undoubtedly more than evaluating sentences for their

meaningfulness. Even if such an evaluation is important to conceiving, it is incumbent upon us to

discover what the other part of conceiving is. The terms 'round' and 'square' are both meaningful terms.

But I cannot fathom what it would be to conceive of a round square's existing.


Option 2 would not raise an objection to the thesis that we cannot conceive of the existence of

impossibilities. Rather, it would be an objection to particular claims about what we cannot conceive.

So, it might be that we can conceive of some round-square's existence and we think we can't because

we are confused and their existence isn't impossible at all. But, even if something like this were true, we

could still hold that whatever is genuinely impossible is inconceivable. So long as we admit that we may

be confused regarding what is impossible, this option should not pose any threat to our opening thesis

about the extent of what we can propositionally conceive, i.e. that it is limited to what is possible.


Option 3 is probably not an entirely legitimate option. What I mean is that if one sincerely

makes a claim to have conceived something and he is familiar with conceiving, he must be conceiving

something, even if what is being conceived is not what he thought it was.


This brings us to option 4. Option 4 is the option with the most intuitive pull. We can imagine

that someone might picture something that shifts shapes from circular to square and describe it as a

round square. And the existence of that shape-shifting object is not hard to conceive. So, option 4 is at

least plausible in some cases. And, it is undoubtedly true that we do make mistaken reports about not










There is a strong pull to add the auxiliary 'can' to 'conceive that' as in (c):


(c) I can conceive that George is a minister.


(c) sounds natural and may convey what we normally want to convey with (a) but there is still an

important semantic difference between 'can conceive that' and 'conceive that'. One may truthfully and

sincerely assert that he can conceive that p without having conceived that p. Normally, when we make

an assertion that we can conceive that p, it is assumed that the evidence one has that he can conceive

that p is that he has conceived that p. Although it may be difficult to imagine what evidence we could

have for our being able to conceive that p without conceiving that p, it is not semantically required that

one conceive in order to truthfully and sincerely make an assertion that one can conceive that p. If such

a requirement were to be made then (d) should not make sense to us:


(d) John can conceive that George is a minister but he hasn't yet conceived that George is

a minister.


When we put the assertion in the third person it is easier to separate the act of conceiving from the

ability to conceive. Our evidence for thinking (d) true is different from the reason we would have for

thinking (c) true. While we would normally have as our justification for believing (c) the fact that we had

performed some act of conceiving, our justification for believing (d) would be some belief we have

about John's abilities. So the 'can conceive' locution does not merely convey something about

performing an act but rather something about the ability to perform such an act, whether or not one

has in fact performed the act. So, I will not in this study use the 'can conceive' locution. Even if it is

normally used to report the performance of an act, we should use some other locution to keep the

other semanticallyy correct) use of 'can conceive' from running interference.


I will instead use the auxiliary verb 'have' and put the reports in past tense, as in (e):










of staircases that face two directions at once. What is particularly insidious about this type of confusion

is that there are no more concepts or physical laws that one must grasp or know in order to detect the

incoherence.


I do not here want to argue that this is the type of confusion that is going on when people

allegedly conceive that there be zombies. I only want to make the more modest claim that this is a

possible confusion that should be explored.


Some confusions of this type are going to be harder to fall into than others. Few people will fall

into a confusion as to whether they can conceive there to be round squares. Almost no one will think

that he or she can. The fact that some forms of this type of confusion rarely occur, e.g. rarely will

anyone think he can conceive that a round square exists because of a confusion between conceiving-of

and conceiving-that, should not prejudice us toward thinking that this type of confusion does not occur

in other instances. When we conceive of round squares we do not tend to confuse that act with

conceiving there to be round squares. The reason we do not is that in order to conceive of round

squares we must grasp both the concepts round and square. And the concepts round and square are

such that grasping them immediately puts one in a position to know that the two concepts cannot both

apply to the same object simultaneously. But there are some concepts, unlike round square, which are

incoherent but whose incoherence is not evident simply in virtue of grasping the concept or concepts

involved.


Consider a concept like greatest prime number. One might grasp greatest prime number and

not detect the incoherence. We can imagine a competent speaker sincerely reporting that he has

conceived that there be a greatest prime number and that he grasps greatest prime number. Either the

speaker has failed to engage in enough rational reflection to conclude that there is no greatest prime

number or he has not engaged in that evaluative type of reflection. In other words, he has either prima










3. It is possible that P be true and M be false. (2,1)


4. If P entails M then it is not possible that P be true and M be false.


5. ~(P entails M) (4, 3)


6. Physicalism is false.


Even though the argument can be given without talk of possible worlds, I believe that the

heuristic of possible worlds is useful for this discussion. If we keep in mind that it is used only as a

heuristic, there should be no problem. It is enough that the argument can be made without the use of

the heuristic to show that one need not be committed to possible worlds in any substantive way to

argue against physicalism by way of the conceivability of zombies.


It is controversial whether zombies are conceivable. Physicalists often argue that the existence

of zombies is inconceivable and that it only seems conceivable. We do not have all of the relevant

physical information regarding what it would take for zombies to exist. One might prima facie conceive

some proposition P which is in fact ideally conceivable, but that would not provide strong enough

evidence for the claim that P is possible. So, while we may prima facie conceive that zombies exist, we

cannot conceive that they exist upon ideal rational reflection. And so, the argument against physicalism

fails.


I would like to suggest that there may be another type of confusion at work. If the type of

conceiving one is engaged in when he allegedly conceives that zombies exist is not propositional

(conceiving-that) then the argument against physicalism fails. The first version of the argument above

turns on premises 5 and 6. If the type of confusion I suggest were to occur then premise 5 would look

like 5':










other falling under a concept, e.g. that of a tree, and that there be things like that which are not

conceived. Whether or not Berkeley made this mistake, it is a mistake that one might make on his way

to arguing for phenomenalism.


In this section, I have discussed a way in which one might be confused about the limits of

conceiving that would lead one to holding a phenomenalist view. The confusion is about the level of

generality of what we can conceive. Although we can seemingly have conceivings about ordinary

particulars, we cannot in fact.


When we think about ordinary particulars we think about them in terms of some description.

Even if we think about them in terms of some definite description,21 we cannot do so when we conceive.

Conceiving requires us to leave open that either something might in fact be different than we believe it

is or that something could have been different than it is in fact. Even though there may be some

concepts under which ordinary particulars fall necessarily, falling under those concepts is not sufficient

for uniquely picking out some contingent particular. And without such a concept which can only apply

to one contingent particular there is no principled reason that our conceiving about some particular, A,

is not also a conceiving about some other particular, B. But then our conceiving is not about A or B but

rather about the application conditions of the relevant concepts that both A and B fall under.


One might come to hold a phenomenalist view if one were to be confused in the way just

discussed. If one were to think that conceiving could be about particulars then he would come to the

conclusion that when one attempts to conceive of some unconceived object, he has conceived of some

particular and thus the object in question is no longer unconceived. And so it will go with all particulars.

But I am suggesting what one conceives of must be at a higher level of generality. If what is conceived is




21Although we often think about particulars through some indefinite description.







































































81










4.3 The Implications of the Distinction for Zombies


Physicalism is a thesis in philosophy of mind about what types of things there are in the world.

According to physicalism, all things in our world, including minds, are wholly determined by the physical.

Physicalism does not hold that necessarily everything is physical or determined by the physical. It is

rather a supervenience thesis, the thesis that all the properties of things are supervenient on physical

properties. The idea is, roughly, that fixing the physical, everything else follows. The purpose of putting

physicalism in terms of a supervenience thesis is to guard against independent variation (Jackson, 1994).

That is, physicalism is stated in terms of a supervenience thesis to ensure that the thesis of physicalism

does not allow for mental states to change without a change in physical states. It follows that according

to physicalism there can be no change in mental states without a change in physical states. Frank

Jackson expresses the main thesis of physicalism in this way:


(III) Any world that is a minimal physical duplicate of our world is a duplicate simpliciter of our
world. (Jackson, 1994, p. 164)



A minimal physical duplicate is a duplicate only of the physical stuff of the world and nothing more. The

requirement that we first consider a minimal physical duplicate rather than just a physical duplicate

guards against us considering a physical duplicate of the world that includes ghosts or spirits or any

other non-physical things. It is obvious that if a physical duplicate of our world replete with ghosts and

spirit-stuff were a duplicate simpliciter of our world, this world would have to include non-physical

objects. (III) entails that our world is a minimal physical duplicate of itself. That is, everything in our

world is either a physical object or necessitated by the existence of those physical objects.


There is one contemporary line of argument which relies on the supposition that conceivability

entails possibility to refute (III). In a world which is a minimal physical duplicate of this world there










considering whether a contradiction is entailed. Conceiving-of is not therefore limited to types of

impossible objects, but rather impossible objects are used in the examples to make clearer the

differences between conceiving-of and conceiving-that. Second, reduction ad absurdum is merely a

model for conceiving-of. One must be careful not to confuse the model for the thing being modeled.


Conceiving-of is performed by first entertaining a concept and then drawing conclusions about

application conditions for that concept as with round square above.


There are two important questions that still need to be answered. First, what is conceiving-

that? Second, what is the relation between conceiving-of and conceiving-that? Both questions will be

answered together.


It seems uncontroversial that whatever conceiving-that p is will require one to grasp all of the

concepts in its content. And I have argued that we cannot conceive that p where p entails the existence

of an impossible object. To illustrate, we can sincerely and truthfully assert (CORS) but not (CTRS):


(CORS) I have conceived of a round square. Whatever is a round square must be round.


(CTRS) I have conceived that there be a round square that is round.


One difference between (CORS) and (CTRS) is that truthfully and sincerely uttering (CTRS) requires that

one consider the existence of round squares while (CORS) does not. So we can safely say that

conceiving that p (where p entails the existence of a) requires minimally that one consider a as existing.

The second thing to notice is that one can conceive-of without requiring the attribution of a property to

anything. But conceiving-that ordinarily requires the attribution of some property or properties to some

(type of) object or objects when we conceive ordinary contingent particulars.


That is because conceiving-of requires only a sufficient understanding of concepts and not a

further awareness of whether or not they could apply to any instances. We may even be aware that

64










'conceive that' construction, on at least some uses. Consideration of the difference between the acts

reported using each construction shows that we can conceive of types of impossible objects, which is

not to say that one can conceive of the existence of impossible objects (e.g., one cannot conceive that

round squares exist).


In section 3.2, I consider an objection that might be raised against the view that we can conceive

of impossible entities but not their existence. This objection, raised by Stephen Yablo, though not

directed at the view developed in this study, is that we can conceive both of the existence of God and

the non-existence of God. If, as it has been argued, either the existence or non-existence of God is

impossible, it follows that we can conceive impossible things to exist. I conclude that Yablo's argument

does not constitute an objection to my view. I argue rather that one can explain away the apparent

conceiving of the existence of impossibilities with the 'conceive of'/'conceive that' distinction I am

drawing.


In section 3.3, I propose a reading of Berkeley's argument for phenomenalism. It is, I think, a

plausible reading, but I am not so much concerned with the historical interpretation of Berkeley (or

whether phenomenalism is true) as with using the argument as a foil for the discussion of the limits on

what can be the objects of conceivings-of. I will argue that there are limits to what one can stipulate in

conceiving and that we cannot conceive of contingent particulars as opposed to types. If we can

conceive of particulars we must be able to think about them in some way that uniquely picks them out

as the particulars they are. The starting point for a criterion of identity for ordinary objects is Leibniz's

Law. I will consider whether the criteria of Leibniz's Law provide us with the tools to pick out particulars

in our acts of conceiving. I will conclude that they do not and that in the absence of some way to think

about particulars in our acts of conceiving we should conclude that we cannot perform acts of

conceiving about particulars.

















































For Sara




Full Text

PAGE 1

1 CONCEIVING OF AND CONCEIVING THAT By KEVIN SAVAGE A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2007

PAGE 2

2 2007 Kevin Savage

PAGE 3

3 For Sara

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank Sara, without whose support I could never have stayed motivated. Thanks also go to Kirk Ludwig, whose advice and ability to reason were invaluable. I also thank Gene Witmer, without whom I might not be in philosophy, for his admirable ability and integrity. And lastly, thanks go to Mi chael Ju bien, who very graciously stepped into my committee at the 11th hour.

PAGE 5

5 Table of ContentsACKNOWLEDGMENTS ................................................................................................................................... 4 LIST OF FIGURES ...................................................................................................................................... 6 ABSTRACT ...................................................................................................................................................... 7 1 INTRODUCTION ......................................................................................................................................... 9 2 MOTIVATION AN D IN ITI AL WORRIES ...................................................................................................... 13 2.0 Introduction ............................................................................................................................... ....... 13 2.1 Motivation ......................................................................................................................................... 13 2.2 Ini tial Wo rries for the Project ........................................................................................................... 18 2.3 Conclusion ............................................................................................................................... .......... 23 3 THE REFUTATION OF THREE ASSUMPTIONS ABOUT CONCEIVING ............................................................ 26 3.0 Introduction ............................................................................................................................... ....... 26 3.1 'Conceiving of' and 'Conceiving that' ................................................................................................ 28 3.2 Yablo and Conceiving of the Existence of the Impossible ............................................................... 39 3.3 Confusions That Could Lead to a Berkeleyan View ......................................................................... 47 3.4 Conclusion ........................................................................................................................................ 56 4 THE DISTINCTION BETWEEN CONCEIVING-OF AND CONCEIVING-THAT ........................... 58 4.0 Introduction ....................................................................................................................................... 58 4.1 An Objection to the Distinction Between Conceiving-of and Conceiving-that ................................ 59 4.2 The Relation Between Conceiving-of and Conceiving-that ............................................................. 62 4.3 The Implications of the Distinction for Zombies .............................................................................. 67 4.4 Conclusion ........................................................................................................................................ 74 5 CONCLUSION ....................................................................................................................................... 77 LIST OF RE FERENCES ............................................................................................................................... .... 79 BIOGRAPHICAL SKETCH ............................................................................................................................... 80

PAGE 6

6 LIST OF FIGURES 1 A penrose triangle .................................................................................................................................... 76

PAGE 7

7 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts CONCEIVING OF AND CONCEIVING THAT By Kevin Savage December 2007 Chair:Kirk Ludwig Major: Philosophy Philosophers use technical notions of conceiving to argue for a nu mber of thes es. However, very little has been done to illuminate what this act of conceiving consists in. So, we may fall into one of several traps when relying on what we can conceive in order to argue for some theory or other. For one, we may fail to provide a valid argumen t or we may have assumed our conclusion in one of our founding premises. In order to ensure we are not falling into one of these traps we should gain a fuller understanding of what conceiving is. Once we do that we see that many things that have been assumed about con ceiving are in fact not the case. In particular, philosophers have argued that our ability to conceive that particular things could have been the case shows us that minds are not physical. But this argument rests on the assumption that what we can conceive is always possible.

PAGE 8

8 But there is a type of conceiving, the performing of which does not allow us to draw conclusions about what is possible. That is because there is a type of conceiving which allows us to conceive impossible objects. And so, if we are not careful we may believe we are perform i n g an act of conceiving relevant to demonstrating the possibility of what we conceive but turn out to be wrong in fact. Thus, our arguments which rely on what we can conceive would fail in those cases.

PAGE 9

9 CHAPTER 1 INTRODUCTION The fact that something is conceivable or inconceivable has been used as the basis for many philosophical arguments. For example, the conceivability of various scenarios has been used to argue for skeptical theses. The supposed inconceivability of an object being unconceived was used by Berkeley to argue for phenomenalism. Contemporarily, the conce ivability of so called philos ophical zombies has been used to argue against physicalism. The claim of the conceivability of any of the relevant situations relies on our having conceived those situations.So, there is some act we must perform, the performing of which allows us to hold that the si tuation is conceivable. Therefore, all of the afo rementioned arguments, as well as others, rely on our being able to perform this act of conceiving. But, if we do not understand what it is we are doing when we conceive, we may not be able to recognize when we have successfully conceived. I will argue in this study that there is a type of conceiving that has not been clearly recognized, but that we do perform, namely, the type of conceiving is normally denoted with reports using the 'conceive of' locution (and its various tenses and moods). There are two comple mentizers 'con ceive' ordinarily takes in assertions: 'of' and 'that'. It has traditionally been assumed that whether 'conceive' takes the complement 'of' or 'that' is inconsequential. Both 'conceive of' and 'conceive that' assertions report the same kind of cognitive operation, according to the tradition. Further, it has been held that both what is impossible is not

PAGE 10

10 conceivable and that we can conceive of particulars (e.g. God, people we know, etc.). In chapter 3, I will argue that all of these assumptions are false. First, I will argue that reports using 'conceive of' and 'conceive that' are reports of substantively different cognitive operations. I will perform a series of comparisons betwe en 'conc eive of' and 'conceive that' sentences in order to show first that 'conceive of' and 'conceive that' are not intersubstitutable and thus not synonymous. I will use further comparisons between sentences of both sorts to demonstrate that there are some 'conceive of' sentences that cannot be paraphrased as 'conc eive that' sentences. This gives us strong reason to think that 'conceive of' and 'conceive that' report different types of operations. The argument that 'conceive of' and 'conceive that' report different operations is intertwined with my denial of the second assumption: that we cannot conceive what is impossible. Second, I will argue that we can conc eiv e of impossible objects, but that we cannot perform any cognitive acts reported with 'conceive that' sentences which are about impossible objects. This is because conceivingthat requires that we think of the objects we are conceiving as existent in some sense. For example, we can conc eive of round squares but we cannot conceive that round squares exist, that there are houses adorned with round squares, or that the president was born in a round square cupola. I will consider an objection not that my requirements for what we can conceive is too weak but that they are too strong. Th e objection supposes that we can conceive that impossible objects exist or conceive of them as existing. I will suggest that what may make it seem conceivable that impossible objects exist is a failure to distinguish between conceivingof and conceiving that. Third, I will argue that we cannot conc eive of particulars or conceive that particulars exist. Of course, particulars do exist. But the way we must think about particulars in acts of conceiving does not uniquely pick out any particulars. Conceiving requires us to leave open that particulars could be or could

PAGE 11

11 have been different than we believe they are or how they are in fact. Anything that we believe picks out or in fact picks out a particular uniquely could be other than how it is or how we believe it to be. In chapter 2, I will discuss preliminary considerations for this study. Section 2.1 will discuss the motivation for the project. In brief, if one either does not understand what conceiving is or does not recognize all of the different ways which we conceive, confusions may result. The specific forms these confusions can take will be looked at in 2.1. Section 2.2 will be an exploration of various prima facie problems for the project. In chapter 3, I will address the three assumptions of the third paragraph. In section 3.1, I argue that there is a substantive distinction between conceivingof and conceiving that. From the observation that 'conceive of' may take as its object an ordinary noun or noun phrase whereas 'conceive that' must take a sentence as its object, I argue that the sentential objects of 'conceive of' may be nouns or nouns phrases which have as their ersatz referents types of impossible objects. In section 3.2, I argue against the suggestion tha t we may conceive that some impossible objects exist (or that there be impossible objects). I suggest that the reason it seems that we can conceive that some impossible objects exist is that we have confused conceivingof for conceiving that. In section 3.3, I suggest a way of understand in g Berkeley's argument for phenomenalism which suggests that Berkeley's mistake might be based on a misunderstanding of the limits of what we can conceive. I argue that one could argue for phenomenalism as Berkeley did if he were to think that we can conceive particulars. Further, I sugg est that this is a mistake an d that wha t appears to be conceiving particulars is conceiving something general. I argue that any time we think we have conceived a particular we have made a mistake because we cannot conceive particulars. In chapter 4, I further delineate the difference between conceiving of and conceiving that while also showing how the two types of conceiving are related and explain why it is important to distinguish

PAGE 12

12 the two. In section 4.1, I consider the objection that all conceiving is conceiving that. I argue that conceiving of and conceiving that are substantively different and that there are common situations in which we conceive of and do not conceivethat. In section 4.2, I suggest that conceiving of is a precondi tion of conceivi ng that. And, in so doing, I attempt to further explain the differences between the two types of conceiving. In section 4.3, I suggest that a failure to recognize the distinction between conceiving of and conceiving that may lead one to the false belief that he has conceived tha t zombi es exist. This, in turn, may lead one to argue that physicalism is false based because it is conceivable that zombies exist (which is required for the zombie conceivability argument against physicalism) though in fact one has only conceived of zombies. I will close the study in chapter 5 with a brief conclusion in which I summarize the findings of the chapters 2 4.

PAGE 13

13 CHAPTER 2 2.0 Introduction In this chapter, I will discuss preliminary considerations for this study, laying out the motivations and addressing some possible problems at the outset. In section 2.1, I discuss more fully the motivation for this study and various problemsthat might arise without a more complete understanding of the act (o r acts ) of conceiving. Those problems are that a lessthan full understanding of conceiving may lead us to make arguments that are either not formally valid or are questionbegging. In section 2.2, I explore the prima facie problems that might be posed for this study. There are two prima facie problems relating to appa rent circularity: the paradox of analysis and what might be called the paradox of performance. There is also a third possible problem which is that one might have to perform an iterated act of conceiving in order to analyze conceive.1I conclude that none of these prima facie problems, whether serious or not, should adversely affect this study. 2.1 Motivation Let me being with looking at a contemporary example of an argument that relies importantly on conceiving to arrive at its conclusion. In Does Conceivability Entail Possibility? (Chalmers, 2002), David Chalmers has argued tha t (the right kind of) conceivability entails possibility. So, if Chalmers is right, if 1Words in italics will be used to denote concepts as in conceive conceiving etc.

PAGE 14

14 one conceives that there be zombies, then it is possible for there to be zombies. And from this possibility, one may conclude that physicalism is false. It is not necessary to go into detail about zombies and physicalism2 to see that the argument will fail if when we think we have conceived that there be zombies, we have failed to do so. The argument will also fail if when we think we have conceived that there be zombies we have conceived zombies in some other way. There is a ge neral con c ern that if one's understanding of conceiving is impoverished in some ways there are at least some arguments that rely on conceiving that one may not make. Suppose someone, Luke, understands that currency is the physical representation of money, that American dollars are the form of currency used in America and that Yen are the form of currency used in Japan Luke mig ht u nderstand what money is in some impoverished sense e.g., that it has physical representations called 'currency', and that the type of currency used is in some way tied to political entities. So, there might be some arguments that rely on graspin g money that Luke could properly make. But were we to know that Luke had this impoverished understanding of money, we would probably not take seriously any arguments Luke makes about, for example, the global economy, nor should we. Similarly, in some instances we should not take seriously arguments relying on conceivi ng made by people who have only an impoverished understanding of conceiving. So, it is important that we get clear about conceiving. If our understanding of conceiving is incomplete, vague, or muddled then we may not be able to make arguments relying on our ability to conceive. And, if our und erstanding is incomplete, vague, or muddled it is unlikely that we would be able to tell when it is that we are justified in making arguments that rely on conceiving. 2This will be done in section 4.3.

PAGE 15

15 To more clearly see the danger, we may examine the structure of the argument that philosophical zombies are metaphysically possible schematically3: 1. ( x) (Conceivable(x) Possible(x)) 2. Conceivable(a) Possible(a) (1) 3. Conceivable(a) 4. Possible(a) (2,3) The argument above is obviously formally valid, assuming univocality of terms. But of course an argument's conclusion may be false even if the argument is formally valid, if one or another premise is not true. The charge ag a inst arguments like Chalmers' that I wish to examine is not that the third premise is false.Normally, one would argue against Chalmers' argument by arguing that the existence of zombies is inconceivable.The worries I want to suggest are a bit different and not aimed only at Chalmers' argum e nt but any argument relying on acts of conceiving. Any argument which relies on some fact(s) about what we can conceive will be adversely affected by an incomplete understanding of the act of conceiving. The first worry is that one might be using different concepts from one step of the argument to the next. Simple misunder standings stemming from the use of homophones or alternate definitions of 3In these examples, 'conceivable' and 'possible' will be treated as predicates, for simplicity's sake. And for present purposes, 'x' and 'a' can take anything whatsoever for their values. These examples are only meant todemonstrate problems that can arise from arguments superficially of this form.

PAGE 16

16 the same word are examples of the type of mistake that might be made. This is the type of mistake that is highlighted by puns.4 Sometimes we make mistakes that are easily discoverable. Suppose that one were to overhear a conversation in which one man, explaining to another why his wife could not make it to the party, says Mary is a little hoarse. A man overhearing the conversation comments to the man with the ill spouse I have an Appaloosa myself. Funny or not, th ese types of mistakes do get made even in situations in which we can give somewhat clear analyses of the concepts in question. We can easily identify the mistake in the example just given. For one, 'horse' and 'hoarse' are spelled differently, so anyone reading an account of th e scenario just describe d, even someone who does not know the difference in meaning between the two, will immediately be predisposed to believe that they 'horse' and 'hoarse' mean different things. And, anyone who does understand what each term means should not have any trouble noticing the differe nce betw een 'horse' and 'hoarse' as the two are so different in meaning. Even if one were to misspell one of the terms in an account above, the difference would still be easily spotted because the contexts of utterance of each term would cue a reader to th e mistak e. But differ ences between what terms mean need not be so obvious. Sometimes differences may go unnoticed. Consider 'unintended' and 'accidental'. Although there is a difference between what the two terms mean, both terms may be used in almost all of the same contexts and they are often used interchangea b ly. If instead of with two terms, what is meant by 'unintended' and 'accidental' were expressed with a single term, it would be incredibly difficult to notice that there were two different concepts at work. And if 'conceive' is a polysemous term which expresses multiple distinct, albeit inter related 4While one does not mistakenly make a pun (in fact, one would not be making a pun if the act in question were unintentional), one plays at making a mistake with a pun.

PAGE 17

17 concepts, then different uses would be particularly difficult to notice. If such a conflation is going on in the argument that it is possible for philosophical zombies to exist, then the argument fails as it is not formally valid.Instead of conforming to the form presented on page three, it wo uld in stead be of th is form: 1. ( x)(Conceivable 1 (x) Possible(x)) 2. Conceivable 1 (a) Possible(a) (1) 3. Conceivable 2 (a) 4. Possible(a) (2,3) The problem then is that there is no valid rule of inference that will allow one to derive the conclusion, 4, from 2 and 3, i.e., the ar gument is no t formally valid. There is a further problem that we might encounter with an argument the success of which depends on the use of an unanalyzed term. We may be in danger of begging the question if, unknown to us, we have smuggled our conclusion into our premises, that is, if one of our premises is in effect the conclusion worded differently. Consider, for example, the terms 'impossible' and 'inconceivable'. Sometimes, when we say that something is inconceivable we mean it is virtually impossible or impossible in the sense of being ruled out by what we know. If at least some senses of 'impossible' and 'inconceivabl e' are synonymous th en it stands to reason that some senses of 'possible' and 'conceivable' are synonymous.If one were to use 'possible' and 'conceivable' in the same sense in an argument like Chalmers' then the argument would be of the following form: 1. ( x)(Possible(x) Possible(x))

PAGE 18

18 2. Possible(a) Possible(a) (1) 3. Possible(a) 4. Possible(a) (2,3) The argument is valid, but questionbegging (and thus not informative), for it assumes the conclusion as a premise. In this section I have tried to motivate the project by explaining several ways in which one may be confused about wha t conceiving is an d why this is important. I have shown that if we are confused in the ways described above, then arguments we take to be valid may not be. A theory of conceiving will help guard against these types of confusions; what we do not know about conceiving can hurt us. 2. 2 Initial Worries for the Project In this section I w ill look at some prima facie proble ms for an examination of conceiving that stem from the nature of conceiving and philosophical methodology. This study will in some instances examine conceiving by means of thought experiments. These thought experiments will sometimes have conceiving and c o nceivability as the object(s) of the thought experiments. They will be used to tell us something about the limits of conceiving and about what conceiving is. There is at least a prima facie worry that might arise about this methodology. I will be examining conceiving and related concepts by way of conceiving. So, there is at least a hint of circularity looming in the wings for my proposed analysis. In section 2.1, I argued that there could be problems with particular arguments which rely on acts of conceiving without a fuller understanding of conceiving I will have to perform the th ought experi ments meant to pro vide a fuller understanding of conceiving without that fuller understanding I

PAGE 19

19 used to motivate the project. So, I will in essence be committing the philosophical sin which the pursuit of this project was to enable us to avoid. But it is not my contention that we have absolutely no starting point. We can perform the act of conceiving without an analysis. But without the an a l ysis we may not always have the tools to determine when it is we have conceived in the way that is relevant to what our aim is. I contend that we perform a type of conceiving that has not yet been discussed in the literature. And, although we do perform this ac t we have not yet examined it so we do not recognize it. If we could not perform any specific mental acts without analyses, we would never have any acts subject to analysis. So, this worry is really a red herring, but I think it important to get it on the tabl e nonetheless. The second worry relates in a general way to the first although more serious. Part of what will be done in this study is both a conceptual and linguistic analysis. But analyses are subject to the paradox of analysis.Below are the definitions of 'conceive' and its forms from the Ox f ord English Dictionary5: Word Part of Speech Definition conceivable adjective that can be conceived of or thought of; imaginable, supposable conceived adjective admitted into, or originated in, the mind; imagined, thought of, etc. conceivement noun conception conceiver noun one who conceives conceiving noun conception 5Definitions are the first relevant definitions for each term in the OED.

PAGE 20

20 conception noun The action or faculty of conceiving in the mind, or of forming an idea or notion of anything; apprehension, imagination. conceive verb to take into, form in, the mind conceiving verb forming in the mind If we look at the list above, 'conceivable' and 'conceiver' both e xplicitly have 'co nceive in their definitions. So on their own, those definitions of 'conceivable' and 'conceiver' will not be helpful in understanding the definition of 'conceive' because we would want a definition to provide us with an explication of 'conceive' that does not have 'conceive' in the definition, for that would be an obviously circular defini tion. If we lo ok at the definition of 'conceive' it is not explicitly circular. While the circularity resulting from the definition of 'conceive' is not precisely the same as the one resulting from particular definitions of 'conceivable' and 'conceiver', it is problematic. What we want to un derstand in trying to un derstand what the act of conceiving is, is not something only linguistic. My contention is only that there is a problem with understanding the term 'conceive' but that there is a problem with understanding what it is to take something into the mind in the relevant way and the ways that we conc eive or tak e something into the mind.6 If 'to take into...the mind' is strictly synonymous with 'conceive', then understanding either term or phrase would require one to grasp precisely the same concept(s). So, in order to grasp conceive one would have to first grasp conceive. This appears to be viciously circular. If in order to understand A one 6Additionally, 'to take into the mind' may be a phrase that describes some of what it is to conceive, but which does not provide necessary or sufficient conditions on conceiving and which is vague. One might describe perceiving as a kind of 'taking into the mind' but it is wrong to think that perceiving is conceiving (alth ough Berkeley did not think so).

PAGE 21

21 must first understand A,then it looks like we might be unable to perform an analysis of A. This would be a problem for any conceptual analysis. However, it is only a problem for analysis if we hold that the analysans must provide us with the ability to grasp a new concept. Th at is not what I hold. Analysis gives us an explanation of a concept we already grasp. We do not gain the ability to grasp specific concepts through analysis. In order to be able to analyze a concept, C, we must be already possess the concept or we would not be able to determine when to appl y C. If one argues to the contrary, he might as well argue that because we come to know something if we correctly analyze knowledge the analysis of knowledge would thereby be circular. Not at all: it simply means the analysis fits the state we are in wit h res pect to it. We cannot require that in order for one to analyze knowledge, he must not prejudice himself toward some specific analysis of knowledge in virtue of what he already thinks knowledge is. Analysis done correctly should provide us with an explicit explanation of what we already grasp. If we perform some act that provides us with something we did not understand going in then either an analysis has not been performed or something has been performed in addition to an analysis. When we perform analysis of some concept, C we consider candidates for application conditions for C and j udge which ones ar e appli cation conditions for C. At least one of the ways we do this is by conceiving of some object or situation, O, and then deciding whether C applies to O. If we decide that C does apply to O, then we move on to some other object or situation in our considerations until we find one to w hich C does not app l y. If C does not apply to O then we hypothesize which characteristic(s) of O would need to be different for C to apply to O.

PAGE 22

22 If we were to analyze conceive in this way, then we would conceive of some act and ask whether that conceived of act is conceiving. If successful, then we would be conceiving of an act of conceiving. Conceiving requires a conceiver. So, a successful analysis of conceive using this methodology would require us to conceive that someone conceive; we mu st rely on an act of itera ted conceiving. That might then require us to first personally experience some mind other than our own. But even hypothetically we can only experience our own mind. So, we might be closed off from this method of analysis. Perhaps we can get around th is worry if we merely acknowledge that the conceiver within the iterated act of conceiving is the same as the conceiver in the first order act of conceiving. So, in order for one to evaluate what it takes to conceive that P, he must conceive tha t he conceives that P. But what it is that I conceive when I conceive that I conceive that P other than conceiving that P is puzzling. Regardless, whether or not I can perform iterated conceivings, iterated conceivings would not tell us anything more about what conceivings are than first order acts of conceiving. But then we do not have the same method of conceptual analysis that we normally do because we would not be conceiving anything that is a candidate for something to which conceiving applies. Rather, the mental act itself would be a candidate for conceiving. I raise this w o rry not because it is particularly problematic but rather to highlight a difference in performing a conceptual analysis of conceive from performing an analysis of some other types of concepts. It is performed by looking at actual (and not hypothetical) acts of conceiving. They may be framed as hypotheti c al acts, particularly when we consider what people are doing when they make specific conceiving reports, but it is our own mental acts that we use as evidence.

PAGE 23

23 2.3 Conclusion In this chapter, I discussed the motivations for this study as well as some prima facie worries that one might have about it. Philosophers have frequently made arguments that rely on conceivability claims. But if we are unclear about what conceiving is or about distinctions between types of conceiving, var i ous problems can arise. Further clarification about conceiving would help us to avoid those pitfalls. But there may be pitfalls in attempting to clarifying what conceiving is. So before the discussion begins it is important to address the worries that might arise at the outset. In section 2.1, I considered two probl e m s that might arise with arguments relying on the application of an unanalyzed concept. First, we might think that there is only one concept we are applying, when in fact we are applying two closely related concepts by way of the same term. If that is the case then argume nts r elying on conceivability cla ims may be invalid if different concepts are used in each claim. Second, if conceivability is also expressed by some other term in the argument that looks as if it expresses something distinct, then some arguments may be in danger of begging the question. For example, Chalmers argues that conceivability entails possibility (and from ther e argues that philosophical zombies are possible), but if conceivability is nothing overandabove possibility then his argument begs the question as the conclusion is not only contained in, but the same in content as one of the premises. In section 2. 2, I discussed three pri m a facie worries for this study. The first and third worries stem from the methodology of conceptual analysis when applied to conceive, conceivable, etc. The second worry is a worry about conceptual analysis in general. The first worry is that if one does not fully understand what conc e iving is, he cannot perform the acts of conceiving necessary for analyzing what conceiving is. However, it was never my contention that

PAGE 24

24 we cannot perform the act of conceiving without analysis.My claim was, less ambitiously, that we need to have a fuller understanding of conceiving in order to be sure that our arguments employing conceiving do not fail to be formally valid or beg the question. So the first worry is not gen uinely a problem for this study The second w orry stems from what is known as the 'paradox of analysis'. We cannot perform any conceptual analysis without already grasping the concept(s) in question. So, analysis cannot bring us any understanding we did not already possess. But this is only a problem if we perform analysis with the aim of ac quiring new concepts or gaining some sort of new understanding. I do not purport to be doing any such thing. Analysis, while not providing us with any new understanding, reveals what state it is we are in with respect to some concept. On e of my aims in chapter three will be to argue that there is a species of conceiving that, while we understand what it is, has not been distinguished from other types of conceiving or has been thought of as something else entirely. Consequently, this study should not be affected by the par a dox of analysis. The third worry is that the method of conceptual analysis might fail in the instance of conceive because we cannot make sense of iterated conceivings. If we were to conceptually analyze book we might conceive of some book candidate and consider whether the concept book applies to the book candida te. If we use the same method when we analyze conceiving, we would conceive of some conceiving candidate then consider whether conceiving applies to it. In order to make sense of this, we must be the conceiver in this iterated act. But whatever we might learn from an iterated act of conceiving ca n be learne d just as well from a first order act. For we can learn just as much through investigation of what we conceive or fail to conceive as we do through considering what hypothetical objects qualify as books

PAGE 25

25 With the motivations and preliminary concerns taken care of, I will in chapter three move onto the argument for the distinction between conceiving of and conceiving that. On the basis of this distinction, I will also be arguing for two other theses. First, I will argue that we can conceive of impossible th ings bu t not that those things exist. Second, I will argue that we cannot conceive about particulars per se. Rather, we can only conceive particulars insofar as they fall under some type.

PAGE 26

26 CHAPTER 3 THE REFUTATION OF THREE ASSUMPTIONS ABOUT CONCEIVING 3.0 Introduction The goal of this chapter is to establish that there are two importantly different types of conceiving which are reported using the locutions 'conceiving that' and 'conceiving of' respectively, the latter of which is not usually distinguished from the former, and to show that conceiving of in the relevant sense is never of particulars bu t only types of things, and that the types of things we can conceive involve inter alia impossible objects. I show in developing this account how this distinction can be used to diagnose some problematic objections to it and some other well known philosophical arguments. The method of this chapte r is to e xamine the sorts of mistakes we make when we conceive or attempt to conceive as a way of articulating what is involved in it. I approach this by examining the types of mistakes people make when they report conceiving something. I consider whether the mistakes people ma ke in these reports are about the content of the act of conceiving or about the act in which they were engaged. Looking at mistakes in conceiving in this way will illuminate both the range of what is conceivable, what conceiving is more generally, and whether there are types of conc e iving heretofore unconsidered. The plan of the chapter is as follows. In section 3.1, I look at the conceive that construction and consider whether it is the construction we should be using exclusively. I argue it is not and in particular that the conceive of construction expresses a s ubstantively different me ntal act from the mental act one reports with the

PAGE 27

27 conceive that construction, on at least some uses. Consideration of the difference between the acts reported using each construction shows that we can conceive of types of impossible objects, which is not to say that one can conceive of the existence of impossible objects (e.g., one cannot conceive that round square s exist). In section 3. 2, I consider an objection that might be raised against the view that we can conceive of impossible entities but not their existence. This objection, raised by Stephen Yablo, though not directed at the view developed in this study, is that we can conceive both of the existence of God and the non existence of Go d. If, as it has been argued, either the existence or non existence of God is impossible, it follows that we can conceive impossible things to exist. I conclude that Yablos argument does not constitute an objection to my view. I argue rather that one ca n ex plain away the apparent conceiving of the existence of impossibilities with the conceive of/conceive that distinction I am drawing. In section 3.3, I propose a reading of Berkeleys argument for phenomenalism. It is, I think, a plausible reading, but I am not so much concerned with the historical interpr etation of Berkeley (or whether phenomenalism is true) as with using the argument as a foil for the discussion of the limits on what can be the objects of conceivings of. I will argue that there are limits to what one can stipulate in conceiving and that we cannot con ceive of contin gent particulars as opposed to types. If we can conceive of particulars we must be able to think about them in some way that uniquely picks them out as the particulars they are. The starting point for a criterion of identity for ordinary objects is Leibnizs Law. I will consider whether th e cri teria of Leibnizs Law provide us with the tools to pick out particulars in our acts of conceiving. I will conclude that they do not and that in the absence of some way to think about particulars in our acts of conceiving we should conclude that we cannot pe rform acts of conceiving about particulars.

PAGE 28

28 3.1 'Conceiving of' and 'Conceiving that' Philosophers tend to be concerned only with the verb 'conceive' when it takes a sentential complement, either a that clause (conceiving that there is extraterrestrial life), a nominalized sentential complement (conceiving there being extraterrestrial life), or an infinitive sentential complement (conceiving there to be extraterrestrial life). They are concerned with the 'conceive that' construction in part be cause there is a tendenc y in philosophy of mind to be concerned with propositional attitudes and in part because of the interest in whether a proposition's conceivability entails its possibility. I will not in this section directly address whether the content of our conceivings must be propositional. Inst e ad, I will consider whether the 'conceives that' construction is the only one that should be considered. It is my view that we should also examine the 'conceive of' construction. The reason that we should look at the 'conceive of' construction is that reports using the 'conceive of' and 'conceive that' construc tions differ not only in the complements 'conceive' takes in those reports but also in that they are reports of substantively different types of acts. If the two constructions are intersubstitutable salva veritate in all reports then 'conceive of' and 'conceive that' are likely synonymous and thus reports using the 'con c eive of' construction would be reports of the same type of act as that expressed in 'conceive that' constructions; the differences would be merely verbal and not substantive. But, if they are not intersubstitutable salva veritate then there is prima facie reason to think that reports with the 'conceive of' or 'con c eive that' construction are reports of different types of acts. However, even if the two constructions are not intersubstitutable salva veritate, the differences between the two constructions may still be merely verbal if one may say everything with the 'conceive

PAGE 29

29 of' construction as one can say with the 'conceive that' construction (and vice versa). But if it can be shown that there are some things that can be said with reports using one construction and not the other, then the two constructions are not used in all cases to report the same acts. Before I ge t to the argum e nt for the distinction between the two types of conceiving, it is important that we first settle on a general form of report that does not prejudice our view of conceiving. If we first look at the present tense uses of 'conceive that' as in (a), it is obvious that there is s o mething unnatural about them: (a) I conceive that George is a minister. (a) reads as if it were a performative like (b): (b) I hereby state that George is a minister. Whatever conceiving is, it is not a speech act, so it cannot be that (a) is a perfor mative, at lea st not in the normal way. Perhaps (a) is a report of some pseudo auditory experience of I conceive that George is a minister the having of which is necessary and sufficient for the truth of (a). But of course one could have a p seudo auditory experience of I conceive that George is a minister without knowing what 'minister' means. So it cannot be sufficient for the truth of (a) that one have a pseudo auditory experience of (a). And while there may be something like a pseudo auditory experience one normally experie nces before an utterance of (a ), one need no t have any, and even if one did it does not seem to have any bearing on the truth or falsity of (a). Since there is this tendency is to read (a) in a way that cannot be the correct way to understa nd (a), we should use some ot her form of the 'conceive that' (and 'conceive of') construction for this discussion.

PAGE 30

30 There is a strong pull to add the auxiliary 'can' to 'conceive that' as in (c): (c) I can conceive that George is a minister. (c) sounds natural and may convey what we normally want to convey with (a) but there is still an important semantic difference between 'can conc eive that and 'conceive that' One may truthfully and sincerely assert that he can conceive that p without having conceived that p. Normally, when we make an assertion that we can conceive that p, it is assumed that the evidence one has that he can conceive that p is that he has conc ei ved that p. Although it may be difficult to imagine what evidence we could have for our being able to conceive that p without conceiving that p, it is not semantically required that one conceive in order to truthfully and sincerely make an assertion that one can conceive that p. If such a requirement were to be made then (d) should not make sense to us: (d) John can conceive that George is a minister but he hasn't yet conceived that George is a minister. When we put the assertion in the third person it is easier to separate th e act of conceiving from the ability to con c eive. Our evidence for thinking (d) true is different from the reason we would have for thinking (c) true. While we would normally have as our justification for believing (c) the fact that we had performed some act of conceiving, our justification for believing (d) would be som e beli ef we have about John's abilities. So the 'can conceive' locution does not merely convey something about performing an act but rather something about the ability to perform such an act, whether or not one has in fact performed the act. So, I will not in this study use the 'c an con c eive' locution. Even if it is normally used to report the performance of an act, we should use some other locution to keep the other (semantically correct) use of 'can conceive' from running interference. I will instead use the auxiliary verb 'have' and pu t th e reports in past tense, as in (e):

PAGE 31

31 (e) I have conceived that George is a minister. (e) is true or false solely in virtue of whether or not one has performed the mental act of conceiving that George is a minister. Now I turn to the difference between those sentences that take the complementizers 'that' and 'of' after 'have conc eived' First, I will look at whether 'have conceived of' and 'have conceived that' are intersubstitutable salva veritate. (e) I have conceived that George is a minister. (f) I have conceived of George is a minister. Obviously, 'conceived of' and 'conceived that' are not intersubstitutable, as evidenced by (e) and (f). While (e ) is truth evaluable, (f) is not a grammatical sentence, and so not truth evaluable. It follows that the two phrases are not synonymous. If there are things that can be said with sentences using the 'conceive of' construction that cannot be said with any sentences usi ng the 'conceive tha t' construction then there is further reason to think that there is a substantive difference in the acts reported by each construction. A mere failure of intersubstitution will not show that there are two substantively different mental acts reported with each locution. So I will attempt to demo nstrate that the way in which 'conceive of' and 'conceive that' fail to be intersubstitutable cannot be bridged by changing other terms in the sentences. First, we should look at whether what is said with (e) can be said with any sentence using 'conceive of'. As seen with (f), 'conceive of' ca nnot be substituted for 'conceive that' in (e). But there

PAGE 32

32 might be some other way to say what is said with (e) with another sentence using the 'conceive of' construction. (g) I have conceived of George being a minister. On a natural reading, (e) seems to be about the utterer having some evidence that George is a minister and also that the utterer cannot rule it out that Geo rge is a minister. On the other hand, (g) is not obviously about some attitude toward George being a minister. In order to make clearer the difference between (e) and (g), we can perform a test. If what is conveyed with (e) is that it is e p istemically open that George is a minister, then it should be contradictory to assert both that one has conceived that George is a minister and that one knows that George is not a minister. So we can test (e) and (g) with (e') and (g'), respectively: (e') I have co nceived that Geo rge is a minister but I am certain (and was certain at the time of conceiving)7 that he isn't (and wasn't) a minister. (g') I have conceived of George being a minister but I am certain (and and was certain at the time of conceiving) that he isn't (and wasn't) a minister. It is clear that (g') is not contradictory, while (e') only seems non contradictory if one forces an unnatural reading. One migh t attempt to alter (e) in some other way than in (e') in order to get the result that what is expressed by (e) is also expressible with some sentence using the 'conceive that' construction: 7The parenthetical clause here is included to guard against a reading to the effect that the knowledge that George isn't a minister came after conceiving that George is a minister. On that reading, the test would be invalid.

PAGE 33

33 (h) I have conceived that George be a minister. If we perform the test with (h) that we did with (e) then we get: (h') I have conceived that George be a minister but I am certain (and was certain at the time of conceiving) that he isn't. (h') is not con tradictory so (h) does not convey that it is episte mical ly open that George is a minister. Rather, (h) seems to be about one's consideration of a hypothetical situation. Further, the utterer of (h) is reporting that he entertained a hypothetical situation in which George was a minister. Whether or not the act rep o rted by (h) is precisely, or only, the consideration of a hypothetical situation is not at issue. The important point here is that (e) and (h) are not reports of the same act. Also, when (h) is compared to (g), it is difficult to see a difference in acts reported by the two sentences. It looks like (g) and (h) repor t the same act. If all 'conceive of' sentences can be used interchangeably with some 'conceive that' sentence as with (g) and (h) then there may not be any substantive difference in the acts reported with sentences using each of the phrases. But so far what has be en suggested by the examples is that what is expressed with all 'conceive of' statements can also be expressed with some 'conceive that' statement'. I want to suggest that there is some act reported by at least some 'conceive of' statements that is substan t ively different from acts reported by 'conceive that' statements. It is no matter to this study that it may not be that there are acts reported with 'conceive that' statements that may not be reportable by any 'conceive of' statement. If that is true, then it would be consistent with th e posi tion that ac ts reported with 'conceive of' are merely a subset of the set of acts reported by 'conceive that'. That would not give evidence that there are two substantively different acts, but merely some that cannot be reported with one of the locutions for what might be merely pra g matic reasons.

PAGE 34

34 What is needed to show that there is a different act reported with 'conceive of' is a demonstration that there are words or phrases that come after the complementizer 'of' that cannot be paraphrased into a 'conceive that' construction. The content of (g) is also expressible as (h). But if we look at (g) th e of clause attributes a property to a particular. But the of clause need not be of that form. If we look at (i),(j), and (k) below, we find that replacing 'of' with 'that' is problematic: (i) I have conceived of George. (j) I have conceived of ministers. (k ) I have conceived of being a minister.8 Clearly, I have conceived that George/ I have conceived that ministers/ I have conceived that being a minister are ungrammatical. One might suggest that (i), (j), and (k) implicitly include that the referent(s) of the ordinary nouns and noun phrases after 'of' exist. So perhaps the content of (i), (j), and (k) may also be expressed by (i'), (j'), and (k'): (i') I have conceived that George exists. (j') I have conceived that ministers exist. (k') I have conceived that being a minister exists. Looking back at (e), it is apparent that when the phrase after 'that' is in the present tense 'conc eive tha t' statements like the three above are about what is epistemically open. But (i), (j), and (k) do not necessarily convey that the existence of anything is epistemically open, even if the utterer of (i), 8There is a reading of (k) which reads the same as I have conceived of my being a minister which would be expressible also with the 'conceive that' locution. I am here assuming that there is a literal reading of (k) that isnot elliptical for I have conceived of my being a minister. It is that reading which is important for the example.

PAGE 35

35 for example, believes that it is epistemically open that George exists.Consider the difference between (l) and (m): (l) I have conceived of Sherlock Holmes. (m) I have conceived that Sherlock Holmes exists. One need not think that it is epistemically open that Sherlock Holmes exists in order to sincerely and truthf ully utter (l) but the belief that it is epistemically open that Sherlock Holmes exists would be required for one to sincerely and truthfully utter (m) as shown by (n) and (o): (n) I have conceived of Sherlock Holmes but I am certain (and was certain at the time of conceivi ng) that he does not exist. (o ) I have conceived that Sherlock Holmes exists but I am certain (and was certain at the time of conceiving) that he does not exist. It is evident that (n) is consistent while (o) is contradictory. Perhaps if the that clause in (o ) is put in the subjunctive mood then the problem can be remedied: (o') I have conceived Sherlock Holmes to exist/that Sherlock Holmes be existent but I am certain (and was certain at the time of conceiving) that he does not exist. There are theoretical reasons to be wary of (o'). First, (o ') treats existence as a property. But existence is not normally treated as a property. Second, even if we table concerns about treating existence as a property, there is a further problem presented for 'conceive that' sentences which are about the existence of fictional characters or types. Some philo sophers hold that fi ctional characters are impossible. Even so, those philosophers can sincerely utter (l). If we look at (p) there does not seem to

PAGE 36

36 be any obvious contradiction and it does not look like people who utter (p) must be irrational in uttering (p) even if they hold that fictional characters are impossible beings: (p) I have conceived of Sherlock Holmes but I am certain (and was certain at the time of conceiving) that it is im po ssible for Holmes to exist. But if we look at (q), there is some significant tension: (q) I have conceived Sherlock Holmes to exist/that Sherlock Holmes be existent but I am certain (and was certain at the time of conceiving) that it is impossible for Holmes to exist. In order to derive a cont r adiction from (q) one would have to hold that conceivability entails possibility. I do not here want to presume that conceivability entails possibility. But even if one does not hold that all fictional characters are impossible there are reasons to think that at least some fictional characters are impossible. An obvious exampl e would be a round square shaped character. Even if we cannot make sense of a round square shaped character, there are other types of fictional characters and creatures we can make sense of that are nonetheless impossible for reasons other than being fictional. In Naming an d Necessity Saul Kripke gave an argument that a unicorn is a type of fictional creature that is impossible for reasons other than just its being fictional.9 Tiger is a species of animal. And unicorn is supposed to be a species in the same way that tiger is. If we found some creature that looked superficially like a tiger but had different organs inside or perhaps different DNA, then such a creature would not be a tiger. The same woul d go for unicorns. But the difference between tigers and unicorns is that while we have some non superficial ways of identifying members of the species, tiger we do not have any non superficial ways of picking out 9Kripke, pp 156 158. Kripke discusses both the metaphysical and epistemological impossibility of unicorns. I have here abridged and combined the metaphysical and epistemological arguments into one.

PAGE 37

37 unicorns. So, were we to come across two creatures resembling horses with horns, each one with vastly different DNA and internal organs, we would have no way of deciding which was the unicorn and which was not. And, the reason we would not be able to tell which is in fac t the uni corn is that neither is. We should consider two claims that might be made about the conceivability of unicorns: (r) I have conceived of unicorns. (s) I have conceived that unicorns exist/ be existent. The preceding digression is important to the discussion because it is not merely that the existence of unicorns is i m possible that should differentiate (r) from (s) but also why the existence of unicorns is (at least according to Kripke) impossible that should help us distinguish the two. The reason that unicorns are allegedly impossible creatures is that there is no description of unicorns that would single them out as members of some species. And, since th ey do not exist, we cannot fix the term 'unicorn' to any particular creatures and those things relevantly like them.So, any attempt to conceive that unicorns exist would have to fail because there would be nothing to distinguish our conc eiving that unicorns exist from our conceiving that something superficially indiscernible from unicorns exists. But then can we conceive of unicorns? I believe that we can. Just think about Kripke's argument that the existence of unicorns is impossible. In order to argue against their existence he must be able to perform some mental act that at le ast resembles conceiving. There must be something towards which his thoughts are directed. And that object is, at least in some sense, the fictional species unicorn. In what way one is able to conceive of unicorns will be addressed later in the chapter. For now, it will be enough that we can in some way conceive of impossible creatures so long as the content

PAGE 38

38 of those conceivings does not entail the existence of the creatures in question, factually or counterfactually. Of course, there is no reason to think that we are able to conceive of impossible creatures or characters and not impossible objects. If we can conceive of unicorns we should also be able to conc eive of round squares. We cannot conceive that round squares exist or conceive of the existence of round squares. But why could we not conceive of round squares simpliciter? It certainly looks like we can in some sense. Again, look at the linguistic evidence. Although we do hold that ro und squares are impossible we also hold th at if they did exist they would be round and square. So, they must in some way be the objects of our thoughts about them. Undoubtedly, the concept round square exists. Conceiving of round squares requires only that one gain an understanding as to what would be required for some object to, per impossible, fall under both r o und and square In this section I have attempted to show that the acts reported by certain 'conceive of' sentences differ substantively from the acts reported by 'conceive that' sentences. First, I demonstrated that 'conceive of' and 'conceive that' are not int ersubstitutable. Paraphrasing that clauses in present declarative tense and mood into present subjunctive will allow 'conceive that' assertions to report many of the same particular acts that are reported by 'conceive of' assertions.But no amount of paraphrasing will make 'conceive that' assertions report what is reported with a 'conc eive of' assertion when the direct object of the complement is a noun or noun phrase, with the paradigm examples being those in which the noun or noun phrase denotes an impossible object.

PAGE 39

393.2 Yablo and Conceiving of the Existence of the Impossible Can we conceive the existence of only what is possible? A complete answer to this question would be beyond the scope of this paper. Regardless, most accept that there is at least some connection between conceivability and possibility. Conceiving seems to give us some epistemic access to possibility.10 However, it seems clear is that there are at least some things we do not think can be conceived because they are conceptually impossible. The types of things we exclude are things like the existence of impossible objects and situations which we know are logically impossible. We do not think that anyone can conc eive of th e existence of round squares or dogs that are not mammals. And we think it's inconceivable for a stone to be bigger than itself. But what about people who do think they can conceive of the existence of round squares and stones larger than themselves? That is, how is it that we un derstand what is going on with those people? We have four options: 1. They are genuinely conceiving of some impossibility and we are mistaken that such things cannot be conceived of. 2. Those things we thought were impossible are not impossible. 3. They are not conceiving an ything at all. 4. They are conceiving of something but are mistaken about how to describe what it is they are conceiving. 10Whether or not conceiving is a reliable guide to possibility is another matter.

PAGE 40

40 If there is any limit to what is conceivable, it is most likely the existence of impossible objects. So, if option 1 is correct then there is likely no meaningful string of words that does not correspond to something that can be conceived. In itself, this is not an obje ction but an observation that should make one uneasy about going with option 1. It is conceivable that P would be nearly synonymous with 'P' is a meaningful sentence. Conceiving is undoubtedly more than evaluating sentences for their meaningfulness.Even if such an evaluation is important to conceiving, it is incumbent upon us to discover wha t the other part of con ceiving is. The terms 'round' and 'square' are both meaningful terms. But I cannot fathom what it would be to conceive of a round square's existing. Option 2 would not raise an objection to the thesis that we cannot conceive of the existence of impossibilitie s. Rather, it would be an objection to particular claims about what we cannot conceive. So, it might be that we can conceive of some round square's existence and we think we can't because we are confused and their existence isn't impossible at all. But, even if something like this we re true, we could still hold that whatever is genuinely impossible is inconceivable. So long as we admit that we may be confused regarding what is impossible, this option should not pose any threat to our opening thesis about the extent of what we can propositionally conceive, i.e. that it is li mite d to what is possible. Option 3 is probably not an entirely legitimate option.What I mean is that if one sincerely makes a claim to have conceived something and he is familiar with conceiving, he must be conceiving something, even if what is being conceivedis not what he thought it was. This brin gs us to option 4. Option 4 is th e option with the most intuitive pull. We can imagine that someone might picture something that shifts shapes from circular to square and describe it as a round square. And the existence of that shape shifting object is not hard to co nceive. So, option 4 is at least plausibl e in some cases. And, it is undoubtedly true that we do make mistaken reports about not

PAGE 41

41 just what we conceive but what we see, hear, taste, etc. We are even likely to be mistaken in some instances about our beliefs. We must then wonder what the nature of the mistake is. We are going to assume that the person mistaken has at least a basic understanding of what conc ei ving is so that he does not fail to conceive of anything. Further, we must assume that whoever makes the claim: (RS) I am conceiving of a round square's existence. is a competent speaker of the English language who understands all of the term in question. That is, he does not report RS in virtue of not knowing the mea ning of the term 'round' (or 'square' or 'existence', etc.). If one sincerely asserts RS, is a competent speaker of the English language but fails to conceive of the existence of a round square then he must have attempted to conc eive of ro und square but failed to do so: (MC) For all speakers S and languages L, if S is a competent speaker of L, sincerely utters 'I am conceiving of x', then S sincerely attempts to conceive of x, and if S fails to conceive of x, S succeeds in conceiving of something else. According to the above principle, if one were to report that he conceived of a round square and failed to do so (which I am holding by hypothesis that he must) he would not of course be conceiving of a round square, but he would succeed in concei ving of something. A principle similar to the one above is suggested by Saul Kripke. Stephen Yablo calls a similar view 'Textbook Kripkeanism': A lot of people appear to have drawn the same 'good news bad news' lesson from their reading of Saul Kripke on conceivability. The ba d news is that conceivability evidence, particularly of the 'conceptual' or 'a priori' sort, is highly fallible Very often one finds a statement E conceivable,

PAGE 42

42 when as a matter of fact, E worlds cannot exist. So it is, for instance, with the conceivability of water in the absence of hydrogen, or Hesperus without Phosphorus. The good news is that (although conceivability evidence is fallible) the failures always take a certain form. A thinker who (mistakenly) conceives E as possible is correctly registering the possibility of something and mistaking the possibility of that for the possibility of E There are illusions of possibility, if you like, but no outright delusions or hallucinations.11 Textbook Kripkeanism assumes something about the connection between conceivability and possibility. But one need not assume that conceivability entails possibility to hold onto the spirit of Textbook Kripkeanism. I believe the spirit is captured by MC. Yablo's examination of Textbook Kripkeanism is particularly relevant to the present issue because Yabl o conte nds that one may conceive of impossibilities. Yablo's counterexample to Textbook Kripkeanism centers on the conceivability of a necessary being. The argument goes like this (for some S): 1. Conceiver S can conceive that God exists. 2. S can conceive that God does not exist. 3. If S can conc eive that P then P is possibl e. 4. So, it is possible that Go d exists and also possible that God does not exist. (1, 2, 3) 5. God exists in some possible worlds and does not exist in other possible worlds. (analysis of 4 using the heuristic of possible worlds) 6. But, if Go d e x ists, he is a necessary existent. 11Yablo, Stephen. "Textbook Kripkeanism and the Open Texture of Concepts," Pacific Philosophical Quarterly 81 (2000): p. 108

PAGE 43

43 7. If God does not exist then God is an impossible being. 8. So, if God exists then God exists in every possible world and if God does not exist he exists in no possible world. 9. Textbook Kripkeanism is false. No problem would arise if one were to conceive of the e xistence of Pegasus and also of the non existence of Pegasus.12No impossibilities arise from either the possibility of Pegasus' existence or the possibility of Pegasus' nonexistence; Pegasus will exist in some possible worlds and not exist in others. God is supposed to exist in all of them or none at all. It is impossible that he should exist in some but not others. Yablo's challe nge is to the traditional view that concei vability entails possibility. So, his argument is not directly aimed at a proposal like the one that I am working with currently. But, on his way to arguing against the conceivability possibility entailment, he does suppose that one can con ceive that there be impossible beings or situations. In order for his argument to work it must be either impossible that God exists or impossible that he not exist. Consequently, in order for Yablo's argument to succeed one must be able to conceive of some impossibility. At first glance there are three options: 1. It is inco nceivable that God exists. 2. It is inconceivable that God does not exist. 3. Both are conceivable. 12Or, at the very least, the same problem would not arise.

PAGE 44

44 What I propose is that the problem lies in a confusion between conceiving of and conceiving that. I suggested in the last section that one can conceive of impossible objects but not of the existence of impossible objects. If one conceives that where the subject of the that clause is an i mpossible object, IO, th en the that clause, taken alone, would entail the existence of IO. Likewise, if one conceives of IO's existing then the content of the conceiving also entails IO's existing13 trivially. And it is the existence of IO that is inconceivable, or so I have argued. My proposal would get around the problem that results from Yablo's supposed counterexample to Textbook Kripkeanism because, although there would be some act of conceiving performed when one believes he is conceiving that God exists, it would not be a conceiving that God exists. For, there is a confusion. But the confusion is not about the type of necessity that God possesses or some misunderstanding regarding the other properties God is supposed to possess. Rather, the confusion is about the type of conceiving one is performing. Alth ough 'con ceiving of' is the more natural phrasing for conceiving reports, philosophers almost exclusively use the 'conceiving that' construction. This preference prejudices us toward thinking that all conceiving must be of some propositional content. If we do not presume that acts of conceiving must have propositional content, at least as normally tho ught of, th en we can explain away the apparent problem presented above. What is impossible is that impossible objects exist. It follows (at least intuitively, if not logically) that it is impossible that we can conceive that those impossible objects exist. This is because conceiving that some objects exist requires specifying at least the relevant circu m stances u nder which such an object exists. But it is clear that no such conditions can be specified in the case of impossible objects. So, if God is one of these impossible objects then it is inconceivable that God exists. But I have left it open that one can conceive of God. And so, on e ca n explain away the apparent problem in the argument 13This presumes an intelligible notion of property entailment.

PAGE 45

45 above as a confusion between conceiving of God and conceiving that God exist.There would not be any problem because if one were confused in the way described he would not be both conceiving that God exist and conceiving that God not exist. Rather, he would be conceiving of God and con c ei ving that God not exist. Supposing that God is not an impossible object (and thus exists), the confusion can be explained away in terms of 'conceiving of' and 'conceiving that', but with a slightly different approach. On the model above, what one does (if God exists) when he believes he has conceived that God doe s not exist is conceive of a world without God. If God exists and is metaphysically necessary then a world without God is an impossible object. So while one would be able to conceive of a world without God, he would not be able to conceive that such a worl d exist. Some might try to explain away the apparent conceiving that something impossible exist with only the 'conceiving that' locution. And if that is possible, then one might wonder why we should admit that there is a substantive difference between conceiving of and conceiving that. It looks like David Chalmers already gave us the relevant distinction with his cate gor ies of prima facie and ideal conceivability: Prima Facie Conceivability: S is prima facie conceivable for a subject when S is conceivable for that subject on first appearances. Ideal Conceivability: S is ideally conceivable when S is conceivable on ideal rational reflection. (Chalmers, 2002, p. 147) One migh t think that con ceiving of some object o is the same as prima facie conceiving that o exist and that conceiving that o exist (on my account) is the same as ideally conceiving that o exist (on Chalmers' account). But I hold that conceiving of o need not in clude anything about o 's existence. Consider some

PAGE 46

46 obviously impossible object like the completely red and completely green ball.I do not think we can even prima facie conceive that such an object exist, as even a minimal grasp of the concepts red and green precludes a completely red and completely green ball existing. But, we can conceive of such an object an d dr aw conclusions about what features it would have to have were it possible for it to exist, such as that it would be colored, have a surface, etc. In this section, I have looked at what I suggest are some of the confusions peoplehave when they conc eive. When peoplethink that they have conceived that some impossible object exists at least one of the explanations is that one has confused conceiving of the impossible object for conceiving that the impossible object in question exists (or conceiving it to exist). What one can conceive as existi ng is limite d by what is possible. But what one can conceive of is not so limited. I then considered an objection not to the suggestion that one can conceive of impossibilities but to the suggestion that one cannot conceive that impossible objects exist. The objection is one suggested by Stephen Yabl o. Yablo suggests that we can both conceive that God exist and conceive that God not exist. Since God, if he exists, is supposed to be necessary, either his existence or non existence is impossible. I attempted to show that this apparent problem can be explained away with the distinction I suggest between 'con c eiving of' and 'conceiving that'. Lastly, I considered an objection that the kind of solution I suggest has already been offered in a different way. Yablo's confusion could also be explained away with Chalmers' distinction between prima facie and ideal variants of conceiving. And so, one might think that the distinction between c onceiving of and conceiving that is nothing over and above the distinction between prima facie and ideal conceiving. I argued that conceiving of differs in type from prima facie conceiving in that one may conceive of some types of obviously impossible objects which one cannot even prima facie conc eive.

PAGE 47

473.3 Confusions That Could Lead to a Berkeleyan View In this section, I will consider the type of confusion that might lead one to the Berkeleyan conclusion that nothing exists other than minds and mind dependent ideas. Consider the following wellknown passage from Berkeley's Principles of Human Knowledge : But say you, though the ideas themselves do not exist without the mind, yet there may be things like them whereof they are copies or resemblances, which things exist without the mind, in an unthinking substance. I answer, an idea can be like nothing but an idea; a colour or figure can be like nothi n g but another colour or figure If we look but ever so little into our thoughts, we shall find it impossible for us to conceive a likeness except only between our ideas. Again, I ask whether those supposed originals or external things, of which our ideas are the pictures or representations, be the mselves perceivable or no? If they are, then they are ideas, and we have gained our point; but if you say they are not, I appeal to any one whether it be sense, to assert a colour is like something which is invisible; hard or soft, li ke something which is in ta ngible; and so the rest. (Berkeley, I.8) We can only perceive objects by way of their sensible properties. The objects which instantiate sensible properties are either themselves perceivable or are not. If they are, then there is nothing more to objects than their sensible properties. If they are not, then one must suppose that there is some intimate connection between objects and their sensible properties.That intimate connection is supposed to be resemblance. In order for something to resemble a sensible property it must itself be sensible. So, if objects resemble their sensible properties, objects must be nothing over and above their sensible properties. So, there is nothing more to objects than how we perc eive them. Without our perceiving them, then they would not exist. Hence, supposed external objects are mind dependent. Berkeley assumes that the limits our ability to conceive is given by what we per ceive: Hence as it is i mpossible for me to see or feel any thing without an actual sensation of that thing, so it is impossible for me to conceive in my thoughts any sensible thing or object distinct from the sensation or

PAGE 48

48 perception of it (Berkeley, I. 5).14It is impossible to conceive of a sensible object without sensible properties. Clearly, what it is to be a sensible object is to be something which can be perceived. But Berkeley is here confused. The confusion commonly attributed here to Berkeley is that he is confused between the objects of perception and the objects of introspection. I do not want to suggest that Berkeley is not confused in the way people normally suggest he is. What I want to do here is suggest that there is a further mistake Berkeley may be making. The mistake lies in thinking that one ca n con ceive of particulars as such. We normally do act as though conceiving can be about particulars as such. But I want to suggest that that is something we cannot do (some exceptions to be mentioned aside). First, I will examine how holding that our conceiving can be about ordinary particulars could le ad one to pheno menalism. I will then argue that acts of conceiving cannot be about ordinary contingent particulars.15 We should look at an example in which we purportedly do conceive of a particular. Suppose that we try to conceive of the Statue of Liberty. So I conceive of a tall greenish figure on a large pedestal. Conceiving of a tall greenish figure on a large pedestal is unproblematic.16But in order for it to be a conceiving of the Statue of Liberty, we need to provide some way of ensuring that what we conceive of is the Statue of Liberty and not some other tall greenish figure. 14Jonathan Dancy notes that in the 1710 edition, Berkeley continues: In truth the object and sensation are the same thing,and cannot therefore be abstracted from each other. 15Objects like numbers and sets, if they exist, will differ from ordinary contingent particulars in this respect. 16I am here overlooking possible vagueness issues. If terms like 'pedestal' are vague it may be that it is indeterminate whether there are any pedestals.

PAGE 49

49 Perhaps we can provide some definite description of the Statue of Liberty. But for any definite description we give of the Statue of Liberty, that description might have failed to apply to the Statue of Liberty. It might have been shorter, constructed somewhere other than Ellis Island, designed earlier or later than it was, etc. It may be tha t there is no pr oblem with picking out a particular like the Statue of Liberty with some mental acts, but I want to suggest that conceiving is importantly different. When we conceive we are directing ourselves toward modal concerns. So we must leave open tha t particulars could have been different than they are. We might think that the Statue of Liberty falls under the definite description the actual tall greenish statue on Ellis Island. Of course, it cannot fail to be the case that the Statue of Liberty is the actual tall greenish statue on Ellis Island. But when we conceive that the Statue of Liberty be somewhere other than Ellis Island we are considering whether we think that situation could have been actual. But if we represented the Statue of Liberty as the actual tall greenish statue on Ellis Island we could not have conc eived that the Statu e of Liberty actually be assembled anywhere other than Ellis Island. In possible worlds speak, if our modal considerations are about anything, they should be about how the actual world could be, not about how some other world is. Of course, objects like numbers (if they exist) would be exceptio ns, as it is un controversial that they fall under definite descriptions that could not have failed to apply to them. Another notable exception is in first person conceivings. Whenever I conceive of myself first personally, it is undoubtedly me that my conceiving is about. And that is because I do not ne ed to repr esent myself with some definite description in order to guarantee that it is me about whom I am conceiving.17 17Perry, John. The Problem of the Essential Indexical, Nous Vol. 13 No. 1 (1979): pp. 3 21

PAGE 50

50 One might object that when we conceive we need not represent particulars with definite descriptions in order to pick out those particulars. We need only stipulate that it is the Statue of Liberty about which I am conceiving in order for it to be the Statue of Liberty about w hich I am thinking. But one's conceiving of some particular cannot amount merely to stipulating that one has done so. For simplicity's sake, we can consider a quasi perceptual type of conceiving to illustrate that there are limits to what one can stipulate. Suppose that one were to report that he had conceived of a spe c ific shade of blue, say robin's egg blue. We would not think that someone had successfully conceived of robin's egg blue if he had in fact formed a quasi perceptual image of a field of Titian (a shade of red), no matter what one reports or stipulates. Perhaps one migh t object to the example by arguing that quasi perceptual conceiving requires quasi perceptual stipulation. One cannot stipulate merely linguistically. One must have an understanding of the concepts involved in order to stipulate. And, where phenomenal concepts are involved, one must be able to form a quasi perceptual picture of a named color in order to stipulate tha t one will be conceiving of that color. But if that is what is required of stipulating that one has conceived of something, then the stipulation is no different in type from the conceiving. So, it is of no use to argue that one ca n conceive of a specific particular A merely because one stipulates that he conceives of A. One must further have an understanding of A. The problem with conceiving of particulars is that it does not look like we understand what ordinary objects like the Statue of Liberty are in addition to our un derstan ding of the many concepts under which they fall. One might still think that we do not think about particulars as falling under some description. But think about a supposed particular like God. When we think about who God is or would be, the way we think about God re quir es that we think about God as falling under some description. God is that

PAGE 51

51 being which is omnipotent and omniscient.18Anything which meets those criteria is God. Suppose one were to say: (G) I met God but he's not omniscient or omnipotent. A natural response to an utterance of (G) would be that whoever or whatever the utterer met was not God. That is because there are necessary and sufficient condi tions on bei ng God, as there are with numbers.But ordinary particulars have accidental properties. And even if one believes that some ordinary particulars have essential properties, (e.g. such as Humphrey having the property of being human necessarily), it is still controversial whether there are properties are properties the having of which suffices for an object to be the pa rticular object it is. There have been attempts at finding properties that individuate (suffice for being) contingent ordinary particulars, but it is unclear that any of these attempts have been successful. And, even if one had an account of ordinary particulars which included a non controversial sufficient property, it is far from obvious that when we think of ordinary particulars, we think about those types of properties. For example, there have been theories about particulars which have as necessary and sufficient conditions for being some object, O, that O be composed of a certain group of subatomic parti c les. Even if such an accoun t is correct it does not seem like that is the type of thing we think about when we think about ordinary particulars. And, even if we did think about ordinary particulars in that way, we do not have the information regarding which e l ementary particles compose what in order for us to have in mind the relevant properties necessary and sufficient for being any specific complex particular. That is, it may be that one thinks in general that what individuates particulars is something about the particles of which 18There are arguably other descriptions which apply to God, like a being which is omnibenevolent and the being who created the world. But whatever specific descriptions one uses to pick out God is of no matter.

PAGE 52

52 they are composed, but it is unlikely that there are any particulars for which one can enumerate which particles that particular is composed of. So, while some conceivings may be about certain odd particulars like God or numbers, it is because what they are can be captured by a definite description w hich we do in fact have in mind when we think about them. And those definite descriptions which apply to these odd particulars pick out these particulars uniquely and apply to those particulars necessarily. With ordinary particulars, it is not obvious that there are any descriptions which uniquely pick out those particulars and apply to those particulars ne cessarily. Even if there are it is unlikely that we have such descriptions in mind when we think about them. To demonstrate the point that there is no necessary description under which we can uniquely pick out particulars I will look at Max Black's argu me nt against the Principle of the Identity of Indiscernibles. Black's argument takes the form of a dialogue between two characters, A and B. B presents a counterexample to the Principle of the Identity of Indiscernibles (PII): (PII) For all objects x and y and properties F, if (x has F iff y has F) then x = y. (PII) provides a condi tion s ufficing for the identity of particulars. The principle is difficult to state in ordinary English as in (PII)19 but it may be better put in ordinary English if we first put it in the logical equivalent of (PII), (PII2): (PII2) For all objects x and y, if x y then ~for all properties F(x has F iff y has F) 19It is natural to say that for any two objects which have all the same properties, they are identical. But then there is a problem because there is only one object, not two as first stated.

PAGE 53

53 (PII2) may be stated in ordinary English. What (PII2) says is that for any two distinct objects there is at least one property that one has that the other does not. The counterexample presented by B is a world empty except for two brass spheres which have all their intrinsic properties in common. The two sphe res are the same size, same density, same mass, etc. And since they are the only two things in the world, they have all the same relational properties as well. So although there are two objects they do not differ in any properties, which is exactly what (PII2 ) says cannot be. One way to object to Black's counterexample is to propose that space is absolute. So, although there are no purely qualitative properties that the two spheres differ in, they do differ in location. If space is absolute, then Black's counterexample fails but it should not bear on what I am trying to show here I am not arguing that there are no real differences between distinct particulars. I am arguing that there is nothing in our conceivings of particulars that would suffice to determine which particular the conceivings are about. Even if we did include in our conceiving of the two spheres thei r locations in absolute space, it would not be necessary that those two spheres be in the locations that they are; one thing we do seem to know about ordinary particulars is that they could have been in different spatio temporal locations than they are in in fact. Even thoug h there may be some property that one sphere has that the other does not e.g. being at such and such spatio temporal location, it is not a necessary property. We may be able to think about particulars in such a way as to pick them out uniquely But when we conceive we need to leave open that those properties instantiated by those particulars might have failed to be instantiated by those particulars. And we further need to leave open that something else could have had the properties of whatever particulars we are considering.

PAGE 54

54 So then what is it we are doing when we think we are conceiving particulars?This is where Berkeley comes in. What he had right was that when think about objects we think about their properties. Even if there is some underlying substratum, it is not about that that we thi nk. When we conceive abo ut contingent particulars we need to leave open that the particulars instantiate properties other than the ones they instantiate in fact or the ones we believe they do in fact. There is no favored property by which we pick out ordinary particulars, but I am sugg esting that when we supposedly con ceive particulars we are conceiving properties. More precisely, we are grasping a concept or group of concepts. From there we can draw conclusions about what would be required for something to fall under a specific group of concepts all at once. So we conceive of a round square by grasping bo th round an d square. We ca n then draw conclusions about what it would take for something to be a round square. Berkeley does use the 'conceive of' locution that I propose denotes a substantively different act from the act denoted by 'conceive that'. But I want to suggest that Berkel ey is confused as to the level of generality that can be conceived. Berkeley argues that one cannot conceive of an object unconceived. Suppose we have a tree named Joe. If one conceives of Joe, Joe cannot be unconceived. And the same goes for any particular.But the mistake is in thinking that our conceivings ca n be about Joe at all. Wh at one can conceive of is some co instantiation of properties that the tree normally instantiates.20 And what that amounts to is grasping the concepts under which any tree falls.But if all conceiving is at this level of generality, there is no particular that has been conceived at all.While one may not be able to conceive of any particular and have that particular be unconceived, that is only because parti culars cannot be conceived at all. What we can conceive is that there be something or 20One can likewise not conceive that Joe be taller or shorter, etc. but rather conceive that there be a co instantiation of such and such properties.

PAGE 55

55 other falling under a concept, e.g. that of a tree, and that there be things like that which are not conceived. Whether or not Berkeley made this mistake, it is a mistake that one might make on his way to arguing for phenomenalism. In this section, I have discussed a way in which one might be con fusedabout the limits of conceiving that would lead one to holding a phenomenalist view. The confusion is about the level of generality of what we can conceive. Although we can seemingly have conceivings about ordinary particulars, we cannot in fact. When we think about ordinary parti culars we think about them in terms of some description. Even if we think about them in terms of some definite description,21 we cannot do so when we conceive. Conceiving requires us to leave open that either something might in fact be different than we believe it is or that something could have been different than it is in fact. Even though there may be some concepts under which ordinary particulars fall necessarily, falling under those concepts is no t sufficient for uniquely picking out some contingent particular. And without such a concept which can only apply to one contingent particular there is no principled reason that our conceiving about some particular, A, is not also a conceiving about some other particular, B. But then our con c eiving is not about A or B but rather about the application conditions of the relevant concepts that both A and B fall under. One might come to hold a phenomenalist view if one were to be confused in the way just discussed. If one were to think that conceiving coul d be about particulars then he would come to the conclusion that when one attempts to conceive of some unconceived object, he has conceived of some particular and thus the object in question is no longer unconceived. And so it will go with all particulars. But I am suggesting what one con c eives of must be at a higher level of generality. If what is conceived is 21Although we often think about particulars through some indefinite description.

PAGE 56

56 at a higher level of generality then there is no conceptual impossibility as there is with some specific particular being both conceived and unconceived. 3.4 Conclusion In this chapter I have looked at some mistakes that can be made when one allegedly conceives. Each mistake I looked at led to some conclusion about the range of things which can be conceived. In section 3.1, I suggested that a mistake that has been made, particularly in recent dis cussions, is in consideri ng only conceiving about propositional contents. I argued that there are mental acts reported by sentences using the 'conceive of' locution that are substantively different from any acts that can be reported by sentences using the 'conceive that' locution. Those acts that differ are the ones reported by 'conceive of' f o llowed by an ordinary noun or noun phrase. I considered the suggestion that (i)'conceive of A', where 'A' is an ordinary noun or noun phrase, is elliptical for (ii)'conceive that A exist'. I concluded that the (i) is not elliptical for (ii) because we can in some way conceive of i m possible objects (and believe that those objects are impossible). But the way we can conceive of impossible objects must be a way which does not require us to conceive that they exist. In section 3.2, I considered an objection that the existence of impossible objects is conceivable. The obje ction came from an argument made by Stephen Yablo against Textbook Kripkeanism. Yablo's argument presumes that both the existence and nonexistence of God are conceivable. One of either the existence or nonexistence of God is necessary and one of them is impossible. So, if both of them are conceivable, something impossible is conceivable. I a r gued that one might be misled into believing both the existence and nonexistence of God are conceivable if he is confused between conceiving of and conceiving that. Lastly, I argued that in spite of appearances, the distinction between conceiving of and

PAGE 57

57 conceiving that is different from the distinction between prima facie conceiving and ideal conceiving. The difference is made evident by the fact that there are some things that can be conceived of (e.g. round squares) that one cannot even prima facie conceive existing. Prima facie conceiving requires some propositional content which further requir es th at one be able to conceive of conditions on the existence under which those things the content is about exist.But object types which fall under concepts that are obviously conceptually incoherent cannot even be prima facie conceived. That is because some concepts are such that grasping them immediately puts one in a position to re cognize that they can not apply to any objects. In section 3.3, I looked at a confusion one might have that would lead one to have a phenomenalist view. The confusion is about the level of generality at which we can conceive. We cannot conc e ive about specific particulars. What we can conceive about is types of particulars. There is nothing that can be both conceived and unconceived. But that is only a problem for non phenomenalists if we conceive of particulars. But not only do we not have to conceive of specific contingent spatio tempora l parti culars, but we cannot do so. What we do when we purportedly conceive of contingent spatio temporal particulars is evaluate the conditions anything would have to satisfy in order fall under some concept or concepts we associate with those particulars. In chapter 4, I will further explain the act we report with 'conc eiv e of' sentences. I will also further explain the relationship between conceiving of and conceiving that Lastly, I will look at how one might use this distinction to argue that zombie arguments fail.

PAGE 58

58 CHAPTER 4 THE DISTINCTION BETWEEN CONCEIVING-OF AND CONCEIVING-THAT 4.0 Introduction There are three objectives of this chapter. First, I will further defend the distinction between conceiving of and conceiving that. Second, I will discuss the relation between conceiving of and conceiving that. And third, I will suggest how the distinction between these two types of conceiving might help explain away the apparen t conceivability of zombies existing. In chapter 3, I argued for a number of different proposals regarding the range of those things that can be conceived. When philosophers appeal to what can be conceived in arguments, the type of conceiving they normally focus on is reported with the 'conceive that' constr uction (o r similar constructions with a sentential complement). I argued that there is another type of conceiving reported with the 'conceive of' construction. Conceivings of may seem to be about ordinary objects but not be about the existence of those objects. So, one may be able to conceive of some type of thing without conceivi ng th at that type of thing exists. Because we may conceive of types of things without conceiving that they exist, we can conceive of types of impossible objects. I also argued that we cannot conceive anything about particulars qua particulars. Conceiving requires that we leave open the possibility that t hings be different than they are or than we believe them to be. If one conceives that, for example, Smith is taller than six feet tall, his success in conceiving that Smith is taller than six feet tall does not depend on Smith's actual height. In order for it to be Smit h that we have conceive d to be six feet tall, we must be able to differentiate between Smith and, for example, Jones in our acts of conceiving about Smith and Jones. The way in which we pick out Smith is by way of some description(s). But we can conceive that any or all descri ptions which in fact apply to

PAGE 59

59 Smith fail to apply top Smith. There is nothing in the content of our acts of conceiving that would uniquely pick out any particular in all possible circumstances. Particulars exist. I am not promoting an eliminativist thesis. Rather, I hold that the contents of our conceivings represent the world in such a way that th e y cannot be ab out particulars as such. In section 4.1, I will consider an important objection to the view that conceiving of is a substantively different type of conceiving from conceivingthat. Specifically, one might argue that conceiving of, in the instances it differs from conceiving that, is not a type of conceiving at all. I argue that we do recognize this difference in normal parlance and in logic as well, at least implicitly. I conclude that the suggestion that substantively different conceivings of are not conceivings at all is mistaken. In section 4.2, I will atte mpt to show tha t conceiving of is a precondition on conceiving that and further discuss what is required for conceiving that. What is required to conceivethat is more involved than what is minimally required for conceivingof, but one may not conceive that without first conceiving of. In section 4. 3, I will suggest that the fail ure to recognize this distinction could be behind the apparent conceivability of philosophical zombies. A full argument that such a confusion is going on in zombie arguments is beyond the scope of this thesis. However, I will demonstrate how such a confusion could lead one to believe that philosophical zombies are conceivable. 4.1 An Objection to the Distinction Betw een Conceiving-of and Conceiving-that One might argue that there really is no reason to think that in addition to the type of conceiving normally discussed (conceivingthat) that there is also some other type of conceiving (conceivingof). I argued in chapter two that there is a substantive difference between conceivingof and conceiving that which is dem o nstrated by the fact that we do make reports of the form 'I have conceived of x where x is

PAGE 60

60 replaced by a noun or noun phrase. I considered the suggestion that 'I have conceived of x is elliptical for 'I have conceived that x exists ( x to exist)' and found that although the former may seem to be elliptical for the latter for some substitutions for x ', the former is obviously not elliptical for the la tter for all substit utions for x. The substitutions for x which most obviously do not admit of the ellipsis are those substitutions which denote impossible objects. And from that, it should follow that 'conceive of' sentences are in general not elliptical for 'conc eive th at' sentences. One could argue that the problem lies in thinking that we can conceive of impossible objects. Hume made this point in A Treatise of Human Nature : 'Tis an established maxim in metaphysics, that whatever the mind clearly conceives includes the idea of possible existence, or in other wor d s, 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. (Book 1. Part 2.Section 2. Paragraph 8) Hume argues, as many have, that what is impossible is inconceivable. From the passage above,it is clear that Hume's view is that we cannot conceive of conceptual impossibilities.22 The conceptual incoherence of x leads to an inability to conceive of x Whether or not we can form an idea of a golden mountain depends on whether the concept golden mountain is coherent. But merely detecting the coherence of golden mountain is not enough to conceive of a golden mountain on this view. Conceiving of a golden m o untain is something we can do in virtue of golden mountain being a coherent concept. 22Just what types of possibility and impossibility there are is beyond the scope of this paper. For purposes of thisstudy, I am assuming that there are conceptual, logical, and nomological possibilities. Whether conceptual and logical possibility are the same or not is another issue beyond the scope of this paper. However, nothing important should hangon whether or not conceptual and logical possibility are distinct notions.

PAGE 61

61 Round square is not a coherent concept, yet I have argued that we can conceive of round squares. What we cannot do is conceive that round squares exist (conceive round squares to exist). Hume says we cannot clearly conceive of impossibilities. I would take this to be saying that we can not ideally concei ve that there be impossible objects. And that is not in conflict with anything I have suggested in this study. One might still object that even if that is what Hume has in mind, there is no reason to think that there is this distinction between conceiving of and conceiving tha t. But we do conceive of impossible objects at some times, even while admitting that they are impossible. When we talk about round squares we have done something which allows us to sincerely say about them what we do. Even though round square is an incoherent concept, we can make sense of what it would tak e for something to be a round square. We can draw conclusions about what would be true of round squares were they to exist. If we could not conceive of round squares then we should think that (CORS) and (COET) have the same truth conditions: (CORS ) I have conc eived of a round square. (COET) I have conceived of an elliptical triangle. But it is not obvious that (CORS) and (COET) have the same truth conditions. Further, I have the intuition that the truth of (CORS) require an utterer to perform an act different from the act r equired for one to truthfully utter (COET). I think we have this intuition in spite of the fact that we know at the outset that both round squares and elliptical triangles are impossible. In logic, we have a model of this type of conceiving with the reductio ad absurdum. Wh en one performs a reductio, one withholds judgments about the assumption of the reductio until it has been revealed that the assumption implies a contradiction. It is in this way that we conceive of impossibilities. We grasp the concept or concepts and from there draw conclusions. Reductio ad

PAGE 62

62 absurdum is, of course, not a perfect model. Reductio ad absurdum works with propositions but conceiving of manipulates concepts rather than propositions. But if we can make sense of an intuitive notion of conceptual entailment, we see that the two are not far off from each other. Of course, one could object that w h ile we do perform this type of act, it is not a form of conceiving. Instead, it should be called 'considering' or 'supposing'. My reply is that it is fine to call the act whatever one wants to call it. However, as I will suggest in section 4.2, there is an int i mate connection between conceivingof and conceiving that which warrants putting both under the heading of conceiving. And, as I have suggest throughout this study, we do perform both of these acts and use forms of the word 'conceive' to refer to both acts. In section 4.3, I will suggest that the fa ct that we do use 'c onc eive' to refer to both acts may have led to a confusion in contemporary philosophy of mind in a matter of some importance. In this section, I have considered the objection that there really is no distinction between conceivingof and conceiving tha t. I have argued that we do perform something which looks a lot like conceiving of with the reductio ad absurdum. I also argued that we do conceive of impossible objects. The evidence that we do so was that we do think that the contents of conceiving reports of different types of im po ssible objects are different reports. 4.2 The Relation Between Conceiving-of and Conceiving-that I have described conceiving of thus far as an act that puts us in the position to understand the application conditions for some concept, whether it is simple or complex. Because conceiving of does not require that one conclude that the concept is coherent, one may conceive of types of impossible objects. But one can n ot conceive that impossible objects exist.

PAGE 63

63 Before I get to the discussion of conceivingthat and its relation to conceiving of, I will discuss what conceiving of is, which has only been touched upon earlier. In the previous section, I suggested that the reductio ad absurdum provides us with a model of conceiving of. If we examine what goes on in reductios we will get a be tter idea as to what conceiving of comes to. In a reductio we make an assumption in order to show that that assumption, either by itself or along with other premises, entails a contradiction. So, in at least some cases, the assumption itself is necessarily false and so i m possible.23 And in some cases it takes quite a bit of work to reveal the contradiction. That work which reveals the impossibility is analogous to conceiving of. But even if we do reveal that some premise is necessarily false we may draw entailments from it which are not contradictions. Consider (RS): (RS) Round(a ) & Square(a) Although (RS) is necessarily false we can draw inferences from (RS) which are not, such as (R) and (S): (R) Round(a) (S) Square(a) So we can sincerely utter a sentence like I have conceived of a round square and if there could be such a thing it would be round. Conceiving of is analogous to the men tal act that first presents (RS) to us and then gets us from (RS) to (R). Of course, there are limits to the analogy between reductio ad absurdum and conceiving of. The aim of conceivingof is not to draw a contradi ctio n. Conceiving of is an act we perform without 23There can, of course, be reductios within reductios. In some of those cases, there may be assumptions made for purposes of reductio that are not impossible.

PAGE 64

64 considering whether a contradiction is entailed. Conceiving of is not therefore limited to types of impossible objects, but rather impossible objects are used in the examples to make clearer the differences between conceivingof and conceiving that. Second, reductio ad absurdum is merely a model for conceivingof. One must be careful not to confuse th e model for the thing being modeled. Conceiving of is performed by first entertaining a concept and then drawing conclusions about application conditions for that concept as with round square above. There are two important questions that still need to be answered. First, what is conceiving that? Se cond what is the relation between conceivingof and conceiving that?Both questions will be answered together. It seems uncontroversial that whatever conceiving that p is will require one to grasp all of the concepts in its content. And I have argued that we cannot conceive that p where p entails the e xistence of an impossible object. To illustrate, we can sincerely and truthfully assert (CORS) but not (CTRS): (CORS) I have conceived of a round square.Whatever is a round square must be round. (CTRS) I have conceived that there be a round square that is round. One difference be tween (C ORS) and (CTRS) is that truthfully and sincerely uttering (CTRS) requires that one consider the existence of round squares while (CORS) does not. So we can safely say that conceiving that p (where p entails the existence of a) requires minimally that one consider a as existing. The second thing to noti ce is that one can conceive of without re quiring the attribution of a property to anything. But conceiving that ordinarily requires the attribution of some property or properties to some (type of) object or objects when we conceive ordinary contingent particulars. That is because conceiving of requires only a sufficient un derstanding of concepts and not a further awareness of whether or not they could apply to any instances. We may even be aware that

PAGE 65

65 nothing could exist to which the concept round square would apply and still succeed in conceiving of round squares. But once one is aware that round squares cannot exist, he cannot truthfully and sincerely utter something like (CTRS). We should next consider an utterances like (CPE) and (CCRS): (CPE) I have conc eived tha t pigs e xist (pigs to exist). (CCRS) I cannot conceive that round squares exist (round squares to exist). (CPE) is an assertion that one has performed some positive act which includes pigs existing. (CCRS) reports either the failure to perform some positive act including round squares or, more likel y, a detection of some incoherence in the concept round square. At the very least, conceiving that requires that one not detect an incoherence either in the concepts the act of conceiving is about or in the relations that the proposition designated by the that clause puts those concepts in. Chalmers' prima facie conceiving (discussed in se ction 2.2) is a type of conceiving that which may be about impossible objects. Even prima facie conceiving that cannot be about objects that fall under concepts such that grasping them puts one in the position to know they are incoherent. But one can con ceive of conc epts and relations of conce pts even if he knows them to be incoherent at the outset. Conceiving of is an act we perform when we consider hypothetical situations per impossible. There is undoubtedly more to conceiving that than can be explored in this study, and much of that has been explored elsewhere.24 So I will at this point lay out the differences between conceiving of and conceiving that and what the relation is between them. Conceiving of amounts to first entertaining a concept then drawing conclusions about at least some of the application conditions. Conceiving that p 24Much of this exploration has been published in the anthology Conceivability and Possibility (eds. Gendler and Hawthorne).

PAGE 66

66 requires that one already has figured out the application conditions for all the relevant concepts need to evaluate p So, conceiving of is a precondition on conceivingthat. Conceiving that p (where p is about some x ) further requires that one not detect that there are any incoherent concepts in the co nten t of conceiving of x In this section, I have discussed what conceiving of and conceiving that are. Further, I have also discussed the differences between these two types of conceiving and the relation between them. Conceiving of is a two stage process whereby one first entertains a conc ept and th en draws conclusions regarding that concept's application conditions. The figuring out of application conditions is modeled by the reductio ad absurdum. Unlike reductio ad absurdum, the aim of conceiving of is not to draw a contradiction, but rather to draw out whatever inferences one can about the application of the conce pt, coh erent or not. Conceiving that p requires first that one conceive of a, where p is about a. So, conceiving of a is a precondition on conceiving that p. Conceiving that p further requires that one not detect and any incoherence in the concepts under which x falls. So, while conceiv ing of is a pre condition of conceiving that, conceiving that further requires one to evaluate a proposition. If one does not detect that the proposition is necessarily false then one has conceived that p, either ideally or in a prima facie manner.25 If one does detect that the proposition is necessarily false, then one does not conceive that p. 25It is important to note that one will not be excluded from conceiving on the grounds that he has a greater capacity for rational reflection. Even an ideal conceiver can prima facie conceive that p where p is not ideally conceivable. Such a situation would be one in which the conceiver merely abstains from conceiving to the extent his ability allows.

PAGE 67

674.3 The Implications of the Distinction for Zombies Physicalism is a thesis in philosophy of mind about what types of things there are in the world. According to physicalism, all things in our world, including minds, are wholly determined by the physical. Physicalism does not hold that necessarily everything is physical or determined by the physical. It is rather a supe rvenience the sis, the thesis that all the properties of things are supervenient on physical properties. The idea is, roughly, that fixing the physical, everything else follows. The purpose of putting physicalism in terms of a supervenience thesis is to guard against independent variation (Jackson, 1994). That is, physicalism is stated in ter ms of a superven i ence thesis to ensure that the thesis of physicalism does not allow for mental states to change without a change in physical states. It follows that according to physicalism there can be no change in mental states without a change in physical states. Frank Jackson expresses the main th esis of physicalism in this way: (III) Any world that is a minimal physical duplicate of our world is a duplicate simpliciter of our world. (Jackson, 1994, p. 164) A minimal physical duplicate is a duplicate only of the physical stuff of the world and nothing more. The requirem ent that we first consider a mi nimal ph ysical duplicate rather than just a physical duplicate guards against us considering a physical duplicate of the world that includes ghosts or spirits or any other non physical things. It is obvious that if a physical duplicate of our world replete with ghosts and spiritstuff we r e a d uplicate simpliciter of our world, this world would have to include non physical objects. (III) entails that our world is a minimal physical duplicate of itself. That is, everything in our world is either a physical object or necessitated by the existence of those physical objects. There is one contem porary line of argument which relies on the supposition that conceivability entails possibility to refute (III). In a world which is a minimal physical duplicate of this world there

PAGE 68

68 would also be creatures which are minimal physical duplicates of us. Some have argued that it is conceivable for such creatures in such a world to lack mental states. These minimal physical duplicates of us that lack mental states are called 'zombies'. Zombies are also behavioral duplicates of us. They appear to walk. They seem to laugh and cry, but are not amu s ed by what we would think of as jokes and are not saddened by what we would call 'tragedies'. So, a zombie world would be behaviorally and physically just like ours. Although he was not the first to hold that conceivability entails po ssibility (o f some kind) David Chalmers has argued that conceivability entails possibility (Chalmers, 2002). Chalmers is careful to explain that it is only a specific type (or types) of conceivability that entails metaphysical possibility. But all types of conceiving that Chalmers discusses are varieties of conceiving that Jointly holding that zombies are conceivable (in the right way) and that (the right kind of) conceivability entails possibility will commit one to the falsity of physicalism.26 Below, I will give a simplified version of the argument: 1. Any world that is a minimal physical duplicate of our world is a duplicate simpliciter of our world. (thesis of physicalism as in (III)) 2. It is conceivable that there be a a minimal physical duplicate of our world in which th ere are zombies. 3. It is conceivable that there be a minimal physical duplicate of our world which lacks mental 26Chalmers argues that one is actually committed to either the falsity of physicalism or the truth of panprotopsychism. However, it is unclear whether panprotopsychism should be considered a type of physicalism. Regardless, it is not the view that philosophers generally have in mind when they talk about physicalism.

PAGE 69

69 states. 4. This world has mental states. 5. It is conceivable that there be a minimal physical duplicate of our world which is not a duplicate simpliciter of our world. (3, 4) 6. If it is conceivable that there be a minimal physical duplicate of our world which is not a dupli c ate sim pliciter of our world then it is possible that there be a minimal physical duplicate of our world which is not a duplicate simpliciter of our world. (Thesis that conceivability entails possibility) 7. It is possible that there be a minimal physical duplicate of our world which is not a d uplicate simplici ter of our world. (5, 6) 8. There is some world which is a minimal physical duplicate of our world which is not a duplicate simpliciter of our world. (Restatement of 7 through the use of possible worlds semantics) 9. Physicalism is false. (1, 8) One might w orry that this talk of possible worlds tak es them too seriously.I am only using possible worlds here for heuristic purposes and remain neutral on the matter of whether or not they exist. The argument above may also be put in non possible worlds terms: Let P express a proposition that exhausts the physical facts of the world and let M express a proposition that e x hausts the mental facts of the world. If physicalism is true, then P entails M. 1. It is conceivable that there be zombies, i.e. the P be true and M be false. 2. If it is con ceivablethat P be tru e and M be false then it is possible that P be true and M be false.

PAGE 70

70 3. It is possible that P be true and M be false. (2,1) 4. If P entails M then it is not possible that P be true and M be false. 5. ~(P entails M) (4, 3) 6. Physicalism is false. Even though the argument can be given without talk of possible wor l d s, I believe that the heuristic of possible worlds is useful for this discussion. If we keep in mind that it is used only as a heuristic, there should be no problem.It is enough that the argument can be made without the use of the heuristic to show that one nee d not be committed to possible worlds in any substantive way to argue against physicalism by way of the conceivability of zombies. It is controversial whether zombies are conceivable. Physicalists often argue that the existence of zombies is inconceivable and that it only seems conceivable. We do not have all of the relevant physical i nformation regarding what it would take for zombies to exist. One might prima facie conceive some proposition P which is in fact ideally conceivable, but that would not provide strong enough evidence for the claim that P is possible. So, while we may prima facie conceiv e tha t zombies exist, we cannot conceive that they exist upon ideal rational reflection. And so, the argument against physicalism fails. I would like to suggest that there may be another type of confusion at work. If the type of conceiving one is engaged in when he allegedly conceives that zombies exist is not propositional (conceiving that) then the argument against physicalism fails. The first version of the argument above turns on premises 5 and 6. If the type of confusion I suggest were to occur then premise 5 would look like 5':

PAGE 71

71 5'. I can conceive of a minimal physical duplicate of our world which is not a duplicate simpliciter of our world. Premise 6 would remain unchanged: 6. If it is conceivable that there be a minimal physical duplicate of our world which is not a duplicate simpliciter of our world then it is possible that there be a mini mal physic al duplicate of our world which is not a duplicate simpliciter of our world. Whereas the statement in premise 5 was the antecedent of the conditional of premise 6, 5' is not the antecedent of 6. So, thenit does not follow that zom b ie worlds are possible. Likewise, it does not then follow that physicalism is false. The question then is: what is the substantive difference between 5 and 5'? It is the difference between conceivingof and conceiving that has been the main focus of this study. Conceiving of is not evaluative in the way that conceivi ng that is. Conceiving of consists of entertaining a concept (coherent or not) and drawing conclusions as to what falls under the concept, as we do with conceptual analysis. Conceiving that requires that one not detect that the concepts are incoherent, while conceiv ing of does not. Because conceiving of does not require this failure to detect incoherence, one can conceive of impossible objects. Further, since propositions are not the objects of conceivingof nothing one conceives of can be either true or false, so conceiving of cannot have metaphysical implications. So, whether or not zombies are possible, they are conceivable in the se n se that they may be conceived of. To illustrate, one may conceive of fictions.And even if one does not hold that fictions are impossible in virtue of being fictions, there are fictions which are impossible and yet we still do not fail to conc eive in th e way that we conceive other fictions.One might write a story in which all of the physical laws are explicitly stated. Also in this story is a naturally blond haired woman named Callie

PAGE 72

72 whose DNA sequence is also explicitly stated in the story. However, the DNA sequence along with the physical laws entail that Callie be brown haired. Even if we are aware of this problem, we do not have any problem conceiving of Callie and her blond hair. One might argue that wha t we do in such an instance is re ally think of the story as having different physical laws or Callie's DNA being different. But we need not do so. All we need do is ignore the impossibility as in a reductio ad absurdum before we conclude that there is an impossibility. I sugges t that we conceive of zombies in the same way we might conceive of Callie. Further, I want to suggest that the difference between the appearance that zombies are conceivable and zombies in fact being conceivable is not merely a matter of the degree of rational reflection, but rather a confusion between types of conceiving. Consider the penrose triangle in figure 1. A picture of a penrose triangle represents surfaces that go away from and come toward a viewer at the same time. A true penrose triangle is impossible, but we can look at the picture and understand what it depicts. And we can look at the pictur e all at once. But it is not until one considers the picture as representing a way that things could be that a problem arises.When one looks at the picture merely as a picture there is no problem. By this I mean that some impossible thi n gs may be represented with pictures and we do not have any trouble understanding the picture as such. An inconsistent co instantiation of properties may be represented. And, so long as we do not look at the picture as something to be evaluated we do not have any proble m. Of course, th ese pictures are more interesting after evaluating them as pictures of impossible things. But, no matter, we can look at them without evaluating them and understand what they depict. I want to suggest that that is what goes on with zombieconceivings.We may consider what is in some sense a representation of a zombie world. And we even grasp all of the concepts and group them together as one. And in this way we can conceive of zombie worlds just as we can look at pictures

PAGE 73

73 of staircases that face two directions at once. What is particularly insidious about this type of confusion is that there are no more concepts or physical laws that one must grasp or know in order to detect the incoherence. I do not here want to argue that this is the type of co nfusion that is going on wh en people allegedly conceive that there be zombies. I only want to make the more modest claim that this is a possible confusion that should be explored. Some confusions of this type are going to be harder to fall into than others. Few people will fall into a confusi o n as to whether they can conceive there to be round squares. Almost no one will think that he or she can. The fact that some forms of this type of confusion rarely occur, e.g. rarely will anyone think he can conceive that a round square exis ts because of a confusion betwee n conceivingof and conceiving that, should not prejudice us toward thinking that this type of confusion does not occur in other instances. When we conceive of round squares we do not tend to confuse that act with conceiving there to be round squares. The reason we do no t is th at in order to conceive of round squares we must grasp both the concepts round and square. And the concepts round and square are such that grasping them immediately puts one in a position to know that the two concepts cannot both apply to the same object si multaneously. But there are some concepts, unlike round square which are incoherent but whose incoherence is not evident simply in virtue of grasping the concept or concepts involved. Consider a concept like greatest prime number. One might grasp greatest prime number and not detect the incoherence.We can i magine a competent speaker sincerely reporting that he has conceive d that there be a greatest prime number and that he grasps greatest prime number. Either the speaker has failed to engage in enough rational reflection to conclude that there is no greatest prime number or he has not engaged in that evaluative ty pe of reflection. In other words, he has either prima

PAGE 74

74 facie conceived there to be a greatest prime number or merely conceived of a greatest prime number. Whichever is the case, it is important to notice that grasping greatest prime number does not by itself put one in the position to detect the incoherence. In this section, I have looked at what im plica tions the conceiving of/conceivingthat distinction might have for the zombieargument against physicalism. I argued that there could be a confusion between conceivingof and conceiving that in the argument. If there is such a confusion then it would be particularly hard to detect as one may merely conceive of zo mbies and po ssess all of the non inferential knowledge one would need to conceive that zombies exist. But like some fictions and impossible drawings, we can conceive of zombies even if they are impossible. 4.4 Conclusion In this chapter, I have discussed the significance of the distinction between conceivingof and conceiving that. In section 4.1, I considered the objection that the distinction does not really exist. I argued that we do perform an act that we call 'conceiving' that allows us to conceive impossible objects. And conceiving of is that act w hich allows us to conceive impossible objects. In section 4.2, I discussed the relation between conceivingof and conceiving that. Conceiving of is a precondition of conceiving that. I also further discussed the requirements on conceiving that. In order to conceive that p27 one must: (1) first conceive of all x 's that p is about.; and (2) not detect any 27Again, we should keep in mind that one may prima facie conceive that p even if p is contradictory. Hencethe requirement is that we not detect an incoherence rather than there being no incoherence or no detectable coherence, regardless of the amount and quality of rational reflection.

PAGE 75

75 conceptual incoherence in p. There is undoubtedly more to conceiving that but the aim of this section was to give some of the minimal requirements on conceivingthat. In section 4.3, I looked at the zombie argument against physicalism. I argued that one way a physicalist might object to the argument is by pointin g out tha t one who purportedly conceives there to be a world with zombies really is conceiving of zombies. If that confusion is committed by one who alleges to have conceived that zombie worlds exist then the way in which zombies have been conceived does not provide eviden ce that it is pos sible for zombies to exist. And thus, the argument fails to be valid.I further suggested that the way one might conceive of zombies is the way we can conceive of impossible fictions and impossible pictures.We perform these acts by entertaining the appropriate concept, be it simple or complex, and drawing some conclusions which do no t have metaphysical implications.

PAGE 76

76 Figure 1. A Penrose Triangle

PAGE 77

77 CHAPTER 5 CONCLUSION In this study, I have exam in ed the act of conceiving. The main conclusion of this study is that 'conceive of' denotes a substantively different cognitive act from 'conceive that'. This distinction is useful in explaining confusions one might fall into regarding what is conceivable. Before the main arguments for the distinction I motivated the study by s uggesting that without a better understanding of this act, arguments having premises about what we can conceive may prove to be incorrect. There were two main reasons I suggested they may be incorrect. First, if 'conceive' is used to report distinct acts from premise to premise, then argume nts relying on conceivability claims may be invalid. Second, if conceivability is the same concept as some other concept we thought to be distinct, then some arguments relying on conceivability claims may be question begging. I later argued that acts reported by 'conceive of' and 'conceive that' are distinct acts. A failur e to distinguish the two might result not only in problems with validity and question begging but further confusions as well. One confusion one might fall into is believing that it is conceivable that impossible objects exist. I argued that the confusion lies in thinking one has conceived there to be some im possible object. However, what one has in fact done in such cases is conceive of some type of impossible object. Another way one might be confused regards the level of generality about which one may conceive. I argued that conceiving cannot be done at the level of partic u lars. All conceiving is of general types of things. I argued that a confusion about the level of generality about which we can conceive might lead one to mistakenly argue for phenomenalism.

PAGE 78

78 I then considered the objection that conceiving of is not conceiving at all. I concluded that this objection is mistaken. One might call conceiving of by some other name but it has an intimate relation to conceiving that which warrants the use of the term 'conceiving'. When we conceive of we ent e rtain a concept, whether simple or complex, and figure out application conditions for that concept. This type of act is modeled, albeit inexactly, by reductio ad absurdum. In a reductio we perform the operations necessary to draw a contradiction from an assumption. Conceiving of is analogous to the figuring out part of reductio ad a b surdum but is disanalogous in that conceiving of does not aim at drawing a contradiction. When we conceive that p we must know at least some of the application conditions of the concepts in p. And the mechanism by which we come to know these application condi tions is conceivi ng of. So, conceiving of is a precondition on conceiving that. With this distinction, we might be able to diagnose argumentative mistakes we did not earlier believe to be mistakes. Or, we might have a different understanding of the mistakes we do think have been made in argume nts. I suggested th at this distinction might be useful in diagnosing the mistake made when one apparently conceives that zombies exist. Whether or not zombies are conceivable, it is worth exploring whether the mistake I suggest is made with zombie arguments, as with other arguments that rely on conceivability claims.

PAGE 79

79 LIST OF REFERENCES Berkeley, George 1998 A Treatise Concerning the Principles of Human Knowledge edited by Jonathan Dancy (New York: Oxford University Press). Black, Max 1952 The Identity of Indiscernibles in Michael J. Loux (ed.), Metaphysics: Contemporary Readings (New York: Routledge). Chalmers, David 2002 Does Conceivability Entail Possibility? in Tamar Ge ndler and Joh n Hawthorne (eds.) Conceivability and Possibility (New York: Oxford University Press). Hume, David 1978 A Treatise of Human Nature edited by by L. A. Selby Bigge (New York: Oxford University Press). Jackson, Frank 2002 Finding the Mind in the Natural World in David J. Chalmers (ed.), Philosophy of Mi nd: classical and contemporary readings (New York: Oxford University Press). Kripke, Saul 1980 Naming and Necessity (Cambridge, Mass.: Harvard University Press). Perry, John 1979 The Problem of the Essential Indexical, Nous Vol. 13 no. 1, pp. 3 21. Yablo, Stephen 2000 Textbook Kripkeanism and the Open Texture of Concepts, Pacific Philso phical Quarterly 81, pp. 98 122.

PAGE 80

80 BIOGRAPHICAL SKETCH Kevin Savage received his B.A. in philosophy from the University of Florida in December 2003. Kevin is currently a graduate student in the philosophy department at the University of Florida specializing in metaphysics, philosophy of language, and philosophy of mind.

PAGE 81

81