In the last ten years or so, talk of “possible worlds” has become decidedly more fashionable from a logical point of view. And here fashion is well-justified: both from a logical and from a metaphysical point of view, the work of Saul Kripke on the concept of a possible world is as challenging as any contributions since the time of Leibniz himself. It is only fair, then, and indirectly flattering, to complain that Kripke is limiting himself as a major philosopher in ways which Leibniz saw to be essential to balanced philosophy. We confine ourselves in this paper to just two crucial dimensions of philosophical depth when it comes to the concept of possible worlds. On the one hand, Leibniz saw the concept as one which raises very usefully (and which may help to answer) many traditional problems of ethics, including those of free will and theodicy. On the other hand, Leibniz saw that questions about possible worlds may call for a radical approach to sane ontology. Such an approach, “revisionary metaphysics,” systematically reinterprets rather than dogmatically regurgitates the convictions of Plain Men and Theoretical Physicists about—to take a favorite example of Kripke’s—this single table’s being made of many distinct molecules.

According to the good word of Kripke:

“Possible Worlds” are stipulated, not discovered by powerful telescopes. There is no reason why we cannot stipulate that, in talking about what would have happened to Nixon in a certain counterfactual situation, we are talking about what would have happened to him (Naming and Necessity, p. 267).

Often Kripke seems to hold that people and other “objects” that can be referred to by names preserve their essential properties in all possible worlds. And the technical device of rigid designator is used

― 62 ―
to further certain prejudices by building into our very mode of reference to actual and possible objects such beliefs as this: nothing which is conventionally thought not to exist could exist. (See his discussion of “unicorns” in the text and appendix of Naming and Necessity.) This is an ontologically shallow notion of possible worlds; for, as we shall see, essential to it is the belief that possible worlds are characteristically composed of many objects (despite the limiting case of a possible world having only one object) which stand in various “external relations” to one another. Such objects could equally well lie in different counterfactual situations where they would be one among many other objects standing in different external relations. Another presupposition is that our mode of reference, given our conceptual scheme, to one object among other actual and possible objects is crucial for sound thinking about possible worlds. Kripke’s modestly visionary brand of overt essentialism in Naming and Necessity derives from a more banal kind of covert essentialism about ontology. The banal kind tends, we suspect, to kill any deep sense of visionary philosophy. As admirers of the formal benefits to be reaped from Kripke’s approach, we would like to investigate the possibility of extending his approach in such a way as to be able, like Leibniz, to discuss matters of ethics and of radical alternatives in Revisionary Metaphysics.

In order to loosen ourselves up for better work on possible worlds, let us do some straightforward normative ethics for an initial spell. Let us consider the question, “Could a pain be good in some possible world and bad in others?” Afterwards, let us close in on an ontological approach to possible worlds whose acceptance would radically alter many current views about possible worlds.

I

“Could a pain be good in some possible worlds and bad in others?” Would the pain of loss felt by McGovern right after his last election in the actual world become a good pain if it were felt by General Amin for the loss of his actual and much abused power in some other possible world? Have we picked out a possible counterfactual situation? Someone, a certain Luddite called Malwit, would retort that Amin could have only the same kind of pain as McGovern in that counterfactual situation, but not the same pain. As would-be

― 63 ―
Kripkeans, bringing the device of rigid designator to bear, we might say, “Suppose, I, McGovern, have this long, continuing homogeneous pain that I call Oscar and I specifically name it with the rigid designator (Oscar) and wish it on General Amin and imagine his tumbling from power into a long, ghastly acquaintance with Oscar, then I would be wishing it, Oscar, on him in all possible worlds.” It is clear that, as a Luddite, Malwit would not allow this to be a description of a possible world. But then we might leave Kripke far enough to try embracing Central State Identity Theory and say: “Suppose in a possible world a fair little sample of McGovern’s jangling nerves and neurons, the sample identical with Oscar, were implanted at the right place in Amin’s skull and limbs and vibrated as dolorifically as before as part of Amin’s body. Surely a pain that was bad in this world would become a pain that was good in the possible world.” Would Malwit still claim we were talking nonsense? Doubtless he would, but then it turns out that in Malwit’s dialect we can still ask, “Could a pain be good in some possible worlds and bad in others?”

For now, let us take a Kripkean step right back from the contingent identities and other alleged oddities of Central State Materialism. Instead let us try to suppose that if little enough, but quite enough of Amin’s brain and CNS could be transplanted into Smith’s body, then the effects of all this could be that Amin remains Amin and Smith remains Smith, yet Oscar is transferred to Smith. We do not find this hard to suppose or understand. But again Malwit says, “Nonsense! The only relevant effect could be, if personal identities are preserved, that Smith now has the same sort of pain as Oscar.” Finally, let us try to suppose that Malwit’s reply is merely a result of a Luddite but intelligible idiolect and idiosyncrasy, not the upshot of sheer idiocy. So be it. Then the question in his idiolect just collapses boringly, though quite intelligibly, into “Could a specific sort of pain be good in some possible worlds and bad in others?” Thus we have a non-Kripkean criterion of transworld identification for a pain in Malwit’s Luddite way of speaking.

Now if philosophers who wish to be worthy heirs of Leibniz are prepared to be serious about a return to a normative ethics, they must not be afraid to advocate conceptual reform in the course and

― 64 ―
cause of creating a wise analysis. Still less should they fear a philosophy’s traditional business of providing guidance for parents and teachers. For the moment then let us ask questions about the relative value of a pain (or of pain in general) within some different possible worlds, it being clearly understood that these are a very limited subset of possible worlds—those characterized by a pluralist ontology like our familiar one. All pain in general, as well as any particular pains, will necessarily be evil qua pain and its potential occurrences must defeasibly oblige persons to eliminate it, while all pleasure in general and any particular pleasure will necessarily be good as pleasure and defeasibly oblige persons to maintain it. The term “necessarily” partly expresses here a steadfast linking habit, a basic connecting attitude which arises from a soundly humanitarian education. Such an education would aim to inculcate in philosophers two basic ethical truths and to extirpate two famous rivals in bogus epistemology.

TWO ETHICAL TRUTHS (treated by some torn between Kant and Mill like contraries, and treated by some who vaguely hope to reconcile Kant and Mill like subcontraries). (A) When I have good reason to believe that some particular person (myself or another) is in pain, then I have a real, if defeasible, obligation to try to help that person and to try to remove or reduce the pain. (B) We have a real, if defeasible, obligation to try to treat all persons (all rational agents) with respect and fraternal concern as common members of a Kingdom of Ends, thus seeking to offer examples or moral consistency which all other persons can rationally try to follow.

TWO EPISTEMOLOGICAL FALSEHOODS. (C) I know immediately what pain is, a brute sensation, just from having such peculiarly privileged access to my own private experiences that I always know for certain at any time whether I am or am not in pain (cf. Descartes on Clear Ideas). (D) It makes no sense for me to express doubt, certainty, or knowledge concerning whether I am in pain, since my avowals of my own pain are conventionally linguistic forms of pain-behavior within a form of life which determines what can be intelligibly expressed (cf. the later Wittgenstein).

The normative ethic of such a humanitarian education will hold that in any possible world (of the relevant subset) pain, or at least a person’s pain, is in itself undesirable while pleasure (or a person’s

― 65 ―
pleasure) is in itself desirable. Persons owe and are owed effort from others to increase pleasure and decrease pain: to recognize that a person is “an end and not a means” is in good part to accept and respect such a network of reciprocal rights and obligations between persons with regard to pleasure and pain. Fraternal Hedonism, pace Kant, pertains very considerably to Duty and to a moral Kingdom of Ends.

What about our original question “Could a pain be good in some possible worlds and bad in others?” Specifically, what about “Would a pain born of humiliating defeat be a bad pain if it were McGovern’s and good if it were General Amin’s?” Well, the humanitarian answer is that qua pain it is bad whichever person has it and its existence as some person’s pain creates defeasible obligations that the sufferer and others who know of it should relieve it. Someone like R. M. Hare in his essay “Pain and Evil” may reply: “It is a contingent truth that people’s pain in general is bad and it would probably be a contingent falsehood, if Amin were overthrown, that his painful humiliation in that possible world is bad. Though in an imperfect world it may be a good thing.” And the Marquis de Sade might reply: “It is a contingent falsehood that people’s pain in general is bad, and it might be a contingent truth that Amin’s humiliation would be bad. In the best world Thrasymachus and I and other culturalist athletes would be free to enslave and torture.” So, to many, it does not seem to be a necessary truth (regarding any kind of necessity) that a person’s pain, as pain, is bad. Intuitions do differ on that manner. The normative moralist is not afraid to reform language and the humanitarian is not afraid to inculcate a necessity-linkage between Pain and Evil. But the normative moralist does have some weight of present linguistic intuition behind him. In the resulting dialect Moore’s or Hare’s Open Question Method will not work for Pleasure or Pain. And in trying to establish a necessary link between “pain” and “bad,” the humanitarian need not be afraid to reply that in a possible world where I have the power to overthrow Amin my defeasible obligation not to cause him pain is overridden by a right or duty to liberate his victims; that in a possible world where I am the overthrown and rather depressed Amin’s gaoler, my defeasible obligation to cheer him up with dialogue is overridden by a qualified psychiatrist’s warning that a

― 66 ―
megalomaniac like Amin’s chance of regaining a sense of reality depends on my leaving the chitchat to the psychiatrist.

The Cartesian concept of a pain as a Clear Idea or a brute sensation for possibly solipsistic consciousness is intolerable for the possible worlds in humanitarian question. It is an ethically emasculated concept, invoking no ethical correlates but only perceptual judgments. Even the Cartesian who goes no further than the Second Meditation can show some improvement by distinguishing the sense and reference of “pain.” He can say then that he has the concept of a sort of feeling, pain, which will create a network of rights and obligations between him and other persons if there are any. He can say he has this feeling sometimes without walking into a Wittgensteinian trap on the matter of other minds: to understand a “communal concept” (a concept suited for use by possible communities) is not to know how well it applies to the actual world. Wittgenstein’s doctrine that it makes no sense for me to express doubt about or knowledge concerning whether I am in pain gets short shrift from our humanitarian educators. The humanitarians reply that it would be similarly and just more transparently absurd for me to say: “It makes no sense for me to express doubt or knowledge about whether I am not feeling pleasure or not—for when I was young and gurgled pleasurably, my elders taught me first person singular forms of verbally hedonic behavior.” It would be much wiser to quip: “As if a Bay Area husband could not be in very genuine doubt as to whether what he is feeling now, on hearing that his Los Angeles-loving wife accepts his pleas to come home, is really pleasure or not!” More important for humanitarian education in accepting responsibility is getting self-indulgent people (like hypochondriacs among others) to use the word “pain” responsibly; to get them to ask whether they feel a bit of itching and tension, or a slight discomfort, or just a desire for attention, or a mixture, or whether they really do feel pain. The humanitarian wants the self-indulgent man to grasp that it is only pain, something morally noteworthy, which gives him a right to others’ concern; that he had better not gull himself about what counts as pain lest he find himself saddled with too many reciprocal responsibilities to others with a bit of itching and mild tension. The humanitarian equally wants the person with hyper-Spartan temperament or psychological

― 67 ―
disturbance to realize the ability of others to recognize his pain before he does. The humanitarian wants people sometimes to study their inner states observantly, with moral sensitivity. Then they can, as persons, fruitfully entertain doubts and make assertions of certainty or knowledge about their inner states, including pain. And all this will be morally fruitful training for reflection on pains across persons across the relevant possible worlds. And so a humanitarian approach to the concept of pain in possible worlds includes worlds where the conscientious, though formerly self-indulgent, query whether a tooth’s twinge is a pain—morally respecting open texture– –and then say later (as the tooth yammers) “Now I know I am in pain!” Descartes is as mistaken as Wittgenstein here: one excludes dolorous doubt, the other knowledge of the dolor as well.

II

Articulating some parts of a humanitarian ethic for a particular subset of possible worlds takes one a certain distance beyond the banal—some of the best things in Mill and Kant seem reconciled and some of the worst things in Descartes and Wittgenstein seem extruded. Posing ethical questions about possible worlds is also more faithful to Leibniz’ profound efforts to combine theodicy with logic in recharging the whole subject of possible worlds. But the topic of possible worlds needs unmiring in other ways. Kripke tells us in Naming and Necessity that if I use “Nixon” as a rigid designator for that object then I will be talking about it (or him) in all possible worlds—or at least in all where it or he exists. A basic assumption of Kripke’s approach to possible worlds (and of others who invoke such notions as model structures) is that they are either (1) worlds of many distinct individuals (standing, as we used to say, in numberless external relations to one another), or (2) worlds with only one individual to which many more could quite intelligibly be added with the resulting rash of external relations. We would get things off to a better start if we ask questions first about which possible world this actual one might be. It seems to follow from Kripke’s words that to answer such a question we need only point and say “THIS one, the one rigidly designated by the title ‘The Universe’ in our ordinary discourse.” But that would be mere philosophical philistinism. A good deal of the history of metaphysics is the history of efforts to offer

― 68 ―
disparate ontological or “categorial” systems designed to explain the structure underlying what might be called Phenomenal Truth-Checking Conditions (PTCC’s). It is, of course, true that, as humans endowed with similar genetic, sensory, and cultural features, we all share a vast horde of beliefs concerning the “facts.” Whether one is a nominalist or a realist, these genetic and sensory features will allow us to determine the truth of “These two balls are red and heavy.” Or, as we prefer to say, the PTCC’s involved in such mundane sentences are the same for any categorial system. What is different is the relevant set of Transcendental Truth Conditions (TCC’s). The extreme realist might insist that the TCC’s for this sentence involve Twoness and Redness and Heaviness and Ballhood being symplocated; the nominalist that the actual world has nothing to do with such bizarre entities as Ballhood or such weird relations as symplocation. When philosophers start to reason toward the right TTC’s by their intuitions of rightness, the structure sought may be largely the structure of human thought (Kant and Strawson), or largely the structure of ultimate reality (Plato, Descartes, Leibniz, Spinoza, Russell, Bergmann), or a puzzling mixture of both (Goodman and Quine). Some metaphysicians have gone further and insisted that insofar as the world is really the world of mere becoming it is a world of illusion—the actual world, properly described, is one for which, contrary to Kripke’s and Quine’s question-begging uses of notation, the values of properly-bound variables would not be what we assume them to be, even when we try to capture PTCC’s of actual illusions.

But we suggest that quite possibly the most interesting possible world as contender for the actual one is that of rigorous Monism. If we talked about the actual Nixon in a Monist possible world, would we have to talk about him, as Kripke puts it? Well, Yes and No. “He” would be a property. Would he remain an object we would refer to in that world, as Kripke puts it? Well, No, of course. The only possible object would be It, Deus-sive-Natura, The One. Monists are often vilified as being unable to account for the same of range of PTCC’s as Pluralists can. It is claimed (most typically by Russell) that Monism cannot admit of a language expressively stronger than that of ancient (and supposedly moribund) syllogistic. For the Monists are alleged to be confined to subject-predicate

― 69 ―
statements. However, this argument is unsound: the monist is not restricted to “syllogistic reasoning” as Russell understood those words.

The Monist is restricted in other ways, of course, given his TTC’s. When he comes to regiment his language he will, of course, not allow in this regimentation any symbols which (as we would say) appear to refer to separate individuals. The Monist will want his language It-tish, to express Transcendental Truth Conditions, what he takes to be The Correct Ontology. But he wishes to cover the same PTCC’s as are regimentally expressed in the pluralistic languages of Russell, Quine, or Kripke. So we have a condition of adequacy on It-tish: a sentence of It-tish is satisfiable under an interpretation just in case the corresponding pluralist’s sentence is satisfiable under the corresponding interpretation.

Many a realist is convinced that he perceives OBJECTS where the vulgar man perceives properties. Our Monist is simply convinced that PROPERTIES (and PROPERTIES OF PROPERTIES) are to be seen where it is vulgarly supposed that OBJECTS are to be seen. And, of course, our talk of PROPERTIES, like our talk of OBJECTS, collapses into his complicatedly predicative talk about It (Deus-sive-Natura). Even the very quantifiers become predicators. It is sometimes held that any attempt to formulate a Monist language would be a mere waste of symbolic expertise: sheer notational sleight of hand could not affect the way the notation-users saw the world and what they meant about it. This is in part quite right, in a way that shows the limitations of formal logic in philosophy, in part dead wrong in a way that shows why notation is only a small if precious part of the story about possible worlds. Characters who complain that a monist notation and semantics will be pointless are often the very types who just before “explained” that a monist notation and semantics could not even be formulated to fit the “Facts” (PTCC’s). But the fact remains that any language or notation as the vehicle of philosophical vision of a Possible World as the Actual World must be buttressed by advice on acquiring the vision that the language will articulate. Study of the language itself, of the way its Transcendental Truth Conditions represent the PTCC’s differently, when compared to familiar languages is a good start.

― 70 ―

Let us therefore take a look at some sample sentences of more or less regimented Quinean and their more or less regimented versions in It-tish.

1. Something is grey
1.* Whereto It greys, thereto It sometimes/at least once/displays

2. Some man is grey
2.* Whereto It mans, thereto It sometimes greys

3. All men are grey
3.* Whereto It mans, thereto It always greys/greys ever

4. Some man loves some man
4.* Whereto It mans firstly and It mans secondly, thereto It sometimes loves firstly-secondly

5. Riggs loves all women
5.* Wheretoever It womans firstly and It is rigging secondly, thereto
It loves secondly-firstly

In fully regimenting these, the pluralist will treat “some” and “all” as quantifiers, and “Riggs” as a definite description. The monist will regiment the “thereto” clause of 1* as “SOM” and the “always” of 3* and “-ever” of 5* as “OM”; he will also treat ‘rigging’ in 5* as “definite” (but not a description) and indicate this by the “contextually defined” ‘co-shows’ (CS). The sentences would then be fully regimented as

1. (∃x) Gx
1.* WT(G¹, SOM¹)

2. (∃x)(Mx & Gx)
2.* WT((M¹ & G¹), SOM¹)

3. (x)(Mx ⊃ Gx)
3.* WT((M¹ ⊃ G¹), OM¹)

4. (∃x)(∃y)(Mx & My & L(x,y))
4.* WT(WT((M¹ & M² & L^1,2), SOM¹), SOM²)

5. (x)(Wx ⊃ L((y)Ry,x))
5.* WT(W¹ ⊃ WT((L^2,1 & R^2*), CS^2*), OM¹)

We leave the formal details for the appendix, but the intuitive key to giving a semantics for It-tish involves distinguishing what Monists would like to call aspects of It from what they might call modes of It. The aspects of It are merely ‘actions’ It performs, e.g., being grey

― 71 ―
or appearing mannishly or loving firstly-secondly. What the Monist calls a mode we might be tempted to call a set (of aspects). But our metaphysical biases should not be foisted onto the Monist—to him they are not sets of objects of any kind, they are simply modes of It. One mode of It is that of all possible aspects of It. The Monist can tell us about this mode by saying that it is the set containing all atomic expressions of his language. Another mode is that portion of the previous mode consisting of all actual aspects of It. The interpretation of a “k place” atomic expression is merely a mode which is part [= is a subset of] of the kth super-mode [= kth cartesian product] of the mode expressing the actual aspects of It. The interpretation of “SOM” is a mode expressing relationships from the possible aspects to the actual aspects [= a set of functions mapping the first set into the second]. Satisfaction under an interpretation is defined as usual, and all the normal logical notions follow from it. (In the Monist’s explanation to a Pluralist, it may seem that the language is being mapped into itself. Doubtless, the Monist would deny this, and claim to be talking about reality.) Different monistic possible worlds are accounted for by varying the mode which was above called “actual aspects of It.” The PTCC’s for this language and classical quantification theory are identical, and so they are both equally able to “capture the facts.” The difference is in the TTC’s: in the Pluralist’s semantics there is a domain of individuals, and relations are explained as sets of ordered n-tuples; in the Monist’s semantics there are no individuals at all, much less relations holding among them.

So much for looking at the Monist’s language. Another way to understand monism might be to reflect on the form of life that would be exhibited by people in a possible world who see their world as our actual world should be seen. Monism, if it could be humanly domesticated, might prove to be of great moral value: persons might cease to see themselves or their nation as “transcendent,” as “something separate and apart” which is opposed to every “alien” thing, locked in conflict or bitter competition with everyone else.

And as a third way to get into monism, one might consider certain MEDITATIVE TECHNIQUES. We describe here a simple one. The would-be Monist can practice ways of looking at “things” whereby at one moment the “things” seem to be a set of distinct

― 72 ―
objects and then seem to be but one, necessarily indivisible object. I can imagine Lucretius constructing a model of spiky, knobbly atoms described in his De Rerum Natura. I can imagine a huge model of this type which looks to me like a crude sphere covered with separable protrusions that I could easily knock off. I can imagine Lucretius seeing the model as the model of a crude sphere whose numerous protrusions cannot, necessarily cannot, be knocked off: atoms are things that are necessarily indivisible. I can look at a large Jackson Pollock oil painting as a work of art, as a necessarily unbreakable unity: canvas, streaks, gobs, and even frame are not individuals but indissoluble attributes of a perfect individual. With such techniques and other forms of meditation one might well come to see the Monist’s Possible World at least for an intuitive moment as the Actual World.

We respectfully conclude that the notion of Possible Worlds should be approached with more respect for traditional problems of ethics and metaphysics. We recommend that others talk about pains and persons in Possible Worlds with full appreciation and respect for this tradition, and about ontologists’ possible worlds as candidates for Actuality. Traditionally there is more to the notion of Possible Worlds than counterfactual conditionals; and if Possible World talk needs rescuing from an overly narrow view of What Is Possible, so do counterfactuals need saving from stale (or novel) fashions in myopia.

University of Alberta

appendix

A Monistic Predicate Logic

Vocabulary:

F^a,b,…i _n (where n,a,b…i are numerals)

SOMⁱ (where i is a numeral)
v ¬ WT ( ) ,

― 73 ―

Formula:

1. F^a,b,…i _n is a formula
2. If φ and ψ are formulae, so are

¬φ

(φ v ψ)

WT(φ,SOMⁱ)

Some Definitions:

(φ⊃ ψ) =_df (¬φ v ψ)
(φ & ψ) =_df ¬(¬φ v ¬ψ)
WT(φ,OMⁱ = _df ¬WT(¬φ, SOMⁱ)

I^a,b (where a and b are numerals) is a special constant which satisfies both
WT (I^i,i, OMⁱ)
WT (WT(((I^ij & φⁱ) ⊃ φ^j), OMⁱ),OM^j)

The following is a “contextual definition”
WT(φⁱ & ψ^i*, CS^i*) = _df WT(φⁱ & WT((ψ^j⊃ I^i,j),OM^j) & ψⁱ, SOMⁱ)

Conversion of the monist’s language to the pluralist’s language:

c[F^a,b…i] = F_n(x_a,x_b,…x_i) where x_g = x_f iff g = f
c[¬φ] = ¬c[φ]
c[φ v ψ] = c[φ] v c[ψ]
c[WT(φ, SOMⁱ)] = (∃x_i) c[φ]
c[WT(φ,OMⁱ)] = (x_i) c[φ]
c[I^ij] = x_i = x_j
c[WT(φⁱ & ψ^i*,CS)] = c[φⁱ] x_i//( x_i) c[ψ] (free occurrences of x_i in φ are to be replaced by occurrences of ( x_i) c[ψ]

A Monistic Model Structure:

M = <P,U>.

where: P is the (denumerably infinite) “mode” of all possible aspects of It.

― 74 ―

(or, as we would say, the set containing all expressions of the form F^a,b,..i_n and given some ordering)

U is a “sub-mode” of P (the actual aspects of It, nonempty)

Interpretation I in a monistic model structure:

I (SOMⁱ = [h: P → U] —a “mode” of mappings from P into U

I (F^a,b,…i_n) ⊆ U^k —where k is the number of numerals in the superscript. (subsets of relationships that could hold among the actual aspects of It)

Monistic satisfaction (under I in M ):

All formulae pick out certain subsets of the functions which map the possible aspects into the actual aspects; that is, certain subsets of I(SOMⁱ). Let ✓(φ) indicate the set “picked out” by φ.

j(i) = the ith formula of P
g∈✓(F^a,b,…i_n) iff >j(a),j(b),…j(i)< ∈ I(F^a,b,…i_n)
g∈✓(¬φ) iff g∈ (I(SOMⁱ) ~ ✓ (φ))
g∈✓(φ v ψ) iff g∈ (✓(φ)⋃ ✓ (ψ))
g∈✓(WT(φ,SOM^k) iff there is an g∈I(SOMⁱ) which differs from g in at most what they assign to k, and f∈✓(φ)

Truth in a monistic model:

φ is M – logically true iff ✓(φ) = I(SOMⁱ) for every interpretation I
φ is M – logically false iff ✓ (φ) = Λ (empty set) for every interpretation I
φ is M – contingent otherwise

Phenomenal truth-checking conditions:

It is clear that a monistic formula φ is satisfiable in a monistic model under a monistic interpretation just in case c[φ] is satisfiable in a standard model under a standard interpretation. So, whatever the phenomenal truth-checking conditions are for classical logic, they are the same for the monistic logic.

― 75 ―

Moral:

Given any ontological point of view, a language can be constructed which has the same expressive power as classical (pluralist) logic and whose transcendental truth conditions favor that categorial point of view. Hence, ontological questions cannot be decided by logical analysis.