Bergmann, Gustav - Notes On Ontology
Bergmann, Gustav - Notes On Ontology
Bergmann, Gustav - Notes On Ontology
Gustav Bergmann
Nos, Vol. 15, No. 2. (May, 1981), pp. 131-154.
Stable URL:
http://links.jstor.org/sici?sici=0029-4624%28198105%2915%3A2%3C131%3ANOO%3E2.0.CO%3B2-W
Nos is currently published by Blackwell Publishing.
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained
prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in
the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/journals/black.html.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For
more information regarding JSTOR, please contact support@jstor.org.
http://www.jstor.org
Fri May 18 08:18:23 2007
Notes o n Ontology
UNIVERSITY OF IOWA
In a recent book ([9]) Professor Hochberg attempts to derive a contradiction from my assay of acts, first articulated in an essay published
in 1955 ([I]) and then developed through three books ([2], [3], and [4])
during the Fifties and Sixties.' One purpose of this essay, officially the
major one if I may so put it, is to show that the attempt fails. Both
Hochberg's argument and my counterargument will be presented in
the second of the three sections into which the essay is divided. Yet this
middle section is the shortest of the three. The first consists of two long
notes without which the second could not be understood. The last
section consists of a series of seven numbered notes of varying length.
Both the first and the last, particularly the last, also contain some
indications about the major changes that have taken place in my world
(ontology) since, more than a decade ago, I stopped reporting them in
print. For without these indications the deep-lying ground and the
far-ranging ramifications of the present disagreement between
Hochberg and myself would almost certainly be missed. T o provide
some such indications, however ~ u c c i n c tis, ~the other purpose of this
essay. But I shall, as I should, also show that the hard core of my
counterargument does not depend on what happened to my world
during the Seventies.
Natural, improved, and ideal languages. The English we use, also when
thinking and speaking about philosophy, is a natural language. So are
French, German, and so on. The ideal language (IL) so-called, or the
ideal schema, as I would rather call it, is neither a natural language nor
even an artificial one that we could learn to use as if it were a natural
language, except about universes of discourse so limited that learning it
would not be worth our while. The IL is merely a class of diagrams (the
elements of) which some philosophers, including myself, design and,
according to certain rules they also lay down, employ when doing
philosophy. The main features of this device are by now familiar; the
more obvious issues have been thrashed out. So I call attention to only
four such features.
( F l ) Withfour exceptions, of which more i n due course, every primitive
mark of the ZL stands for an existent. Phenomenologically. on
the side of the act, these existents are allsimples. Ontologically, on the side of the intention, they are either things. i.e.,.
either particulars or universals, which are among the determinates; or they are subdeterminates, such as, e.g., the connectives and the quantifiers. The latter I call subdeterminates rather than, as is more usual, subsistents, in order to
emphasize that there are no kinds or degrees of "existence" but only categories of existents, distinguished by
how they combine and by their different degrees of
"separability" and "independence." I call them subdeterminates because every member of this category, so characteristic of my world, is much less separable and independent of determinates than any two or more members of the
latter category among them~elves.~
Adeterminate, finally, is
either a thing or a complex or a class; and a complex is either
a fact or a circumstance.
NOTES ON ONTOLOGY
133
have been, and sometimes still are, used to unpack this use of
'perspicuous'. The only use of the IL is to serve as a standard. The
IL-philosopher will, and practically must, whenever the difference
makes no difference, use mere notations of convenience, i.e., those of
some merely improved language. The IL serves merely asthe standard
by which he decides whether or not the notation he adopts for what he
is about does make a difference for it. Unless he does that, he may get
caught in the snare that nowadays catches so many. He will not, as an
ontologist talk (in Husserl's phrase) about the things but, rather, as a
linguist, about the words people use in thinking and talking about
them. Needless to say, I shall follow my own prescription.
The assay. A simple paradigm will recall the gist of the assay proposed for the act. Let g,(a) be an atomic fact; then an act that (as one
says) intends this fact will be another fact, viz., a second particular
exemplifying (at least) a second and a third simple character, g, and A,,
g, being the thought, or as I first called it. the proposition that (as I
technically use 'intend') intends g,(a), while A, is perceiving, or believing, and so on, depending on what I called and still call an act's species,
or less circumstantially, depending on whether the act is a perceiving, a
believing, and so on.
g,Jtlg,(a) is in the paradigm the connection (not: relation!) between
the thought and "its" intention. g2JtlSl(a)isanalytic, while, for every g3#
g2, g3Jtlg1(a)is contradictory; just as-I resort for the first time to a
semi-schema-every instance of g,Jtl6 except g,Jtlg,(a), is contradictory.
As for the paradigm, so mutatis mutandis for every thought and intention.
The notation I just used and with one exception have always used
is the one Hochberg now also uses. The exception is the notation for
thoughts, deplorably counterstructural and therefore now rightly
criticized by Hochberg, which I have banned from the IL only in the
early Seventies. This improvement, though, makes no difference for
either his argument or my counterargument. So I leave for the second
note what I think should be said about that hapless notation.
Hochberg and I both appeal to a notion of analyticity. The rethinking of this notion I have done during the Seventies produces an
explication of it more complex than any one so far proposed. Yet again,
this development in my world makes no difference for any context
here relevant in which he or I speak of something's being analytic or
contradictory. So we can ignore the later e~plication.~
Hochberg and I both use instances of the schemata d p and a! =
p; we both take them to stand for something's-intending (meaning)something and something's-being-the-same-as-something, respectively. Yet I had better alert the reader at this first occasion to two
interconnected questions we must face. How are the existents, ifany,
NOTES ON ONTOLOGY
135
which the instances of the two schemata stand for, to be assayed? What,
anything, do 'A' and '=' stand for?
When first proposing the assay, I took 'A' to stand for a subdeterminate of the sort to which, in my world, the connectives and the
quantifiers belong. Hochberg still attributes to me this view which I
abandoned during the early Seventies. For argument and counterargument this change, once more, fortunately makes no difference. Yet,
with a phrase from the opening paragraph, if I ignored this particular
change, the import of our present disagreement would only too easily
be missed. So there will be a good deal about the view I now hold about
'A' (not: A)in the second and the third note. The case of sameness is
even more sensitive.
The st.rictest possible notion of sameness, already implicit in my
1955 essay.,first fully articulated in a joint paper with Hochberg published in 19576 and here first mentioned in connection with (F3), is
crucial for our present disagreement. Upon this notion, "two" determinates that are either things or complexes7 are the "same" if and only
if they are one and only one. Or, equivalently on the side of the IL, an
instance of a = /3 is, by (F2) and (F3), analytic or contradictory depending on whether or not the expressions in it to the left and the right of '='
are, as one says, two tokens of the same type; i.e., more explicitly for
once, if and only if these expressions, whether primitive marks or
strings, are mark for mark, including parentheses, if any, different
tokens of the same types in the same order. (As you see I still suppose,
as I said I would until the very end, that the IL is linear and that
parentheses occur in it.) This notion and nothing else I mark by the
instances of a = p. Clearly, it differs radically from the familiar defined
notion I have called Russell-Leibniz identity, marking it by '=R<. Equally
clearly, this older notion, on which we have all cut our teeth, is useful in
a world whose IL is (unramified) PM, with extensionality, supplemented by primitive marks to stand for some things. Its usefulness in
such a world--call it for brevity's sake a PM-world-rests essentially on
the actuality, or if you please, analyticity, of all instances of the schema,
4(a).(a=~~P)>4(P)
The argument, letting the two primitive marks 'a' and 'b' stand for one
particular, takes for granted, or, as I would rather say, it supposes, first,
that a = b is analytic; and it supposes, second, that all the instances of
the (half-) schema 4(a).(a = b) >&b) are analytic. ([9]) Then it considers the instance
NOTES O N ONTOLOGY
137
The objectionable notation writes ,rpl for the thought, if any, whose
intention is p; it makes well-formed all instances of cpJtlp such that the
instances of cp and p stand for a nonrelational character of the first
Russell type and a determinate that is a fact, respectively."
The notation is very convenient, so convenient indeed that I still
use it at my desk whenever a fragment of an improved language will do.
Yet it is objectionable for at least three reasons. First, it is syntactically
counterstructural. Second, it reveals a severe structural weakness of
what was then my ontology. Third, and worst, it encourages by omission
the belief that in that earlier world of mine there are certain existents I
never meant to be there and which, therefore, I should have taken care
to exclude most explicitly. I shall next by my present standards identify
these defects. But do not mistake anything I shall say for a defense of
that notation. All I may show, incidentally, is that with some luck one
may use a confusing notation without being snared by any of the
confusions it invites.
First.In keeping with the paradigm, since 'g,(a)' is a string, so
is ' rg,(a)ll". Yet the formation rules of 1955 and 1957independently of, although of course with a view toward
interpretation-endow ' rg,(a) l i v with all the "syntactical" features
of the marks intended to stand for the members, all simple, of a
single category. In effect, therefore, the marks standing for the
members of a fundamental category are, with a view toward
interpretation, divided into two kinds; those of one are quite
properly primitive, those of the other, most improperly, strings.
That is, to say the least, a case of very bad syntactical manners.
There is still another, although much lesser flaw of the same
kind. Syntactically, the corners are introduced as a new member
of the category some call logical constants. In the interpretation,
however, there is no explicit statement as to what, if anything,
they stand for. This omission, though, is easily accounted for and,
I submit, excusable. on the ground that at that time no member of
that category had in my world achieved ontological status. Implicitly, therefore, the new logical constant is specified as, were it
otherwise defensible, I would still specify it, namely, as standing
for nothing.
Second. Both the 1955 and the 1957 paper distinguish between the formation rules of the calculus and their interpretation.
By virtue of the former ' rgl(a)l Ap1(a)' is well-formed. By virtue
of the latter and because of what the world is like, rgl(a)l Agl(a) is
analytic. That is of course as it ought to be. A cue as to what is
wrong with the way the analyticity is secured is provided by how in
this context the author of the 1955 paper used 'define'. Having
stated the formation rules of the new calculus, to serve as the IL of
his new world, he proceeds to "define" a new notion of analyticity
by adding, ad hoc and (as one says) mechanically, to the collection
of determinates picked out in his old world by his old notion of
analyticity, the collection of all determinates instances of rp7 A@
stand for (and, of course, what becomes deducible from their
"union"). Such ad hoc modification, even if as in this case for a
good reason, diminishes the credibility of an ontology.12 The
thing to do instead is what I have tried to do since, namely, search
for a new explication satisfying more than just a single systematic
need and such that what (in this case) had to be "added" to the
139
NOTES ON ONTOLOGY
extension of the old notion "by definition" falls under the new
one, if I may so put it, a principio et ex principiis.
Third. In 1957 we said that if '. . . . . 'is a well-formed sentence
(of the IL) then ' r. . . . ' will be a well-formed predicate, without
adding, either immediately or at least somewhere in the paper,
the proviso that ' r. . . . .l ' will be admitted into the IL only if there
is in the world at least one act intending (as one says). . . . . . , in
which case ' r. . . . .T ' will stand for this thought. Hochberg now
correctly observes that omitting the proviso may encourage the
belief that in my world there are thoughts unthought. I deplore
this unfortunate byproduct of bad syntactical manners that do
not either acknowledge or keep away from everything not strictly
syntactical. For the rest, I can only say that I have always taken the
proviso for granted and, fortunately, can adduce two items of
evidence for my good faith in saying it. For one, I have always
held and still hold that by theprinciple of Exemplification there is no
universal not exemplified by a particular.13 For another, I have
always rejected the view, attributed to Frege by many students
including myself,14that what he calls thoughts (and concepts) are
existents "outside of minds." And I have, with or without reference to Frege, emphasized this rejection by maintaining that, like
all "mental" universals, the thoughts of my world are exemplified
only by the particulars in acts and therefore "in minds." Thus I
had no structural motive against and two very potent ones for
adopting the proviso.
..
How are the existents the instances of a = pstand for to be assayed; and
what, if anything does '=' stand for? Stay for the moment with the
simples I call things. Nonphilosophers will rarely, if ever, have occasion
to intend some particular's being thesame as itself. Not so for diversity.
We all are quite commonly, at the very phenomenological rock bottom
and even with immediate evidence, presented with one thing's being
diverse from another, universal o r particular. Because of this
phenomenological advantage of diversity ( a # P) over sameness ( a = P)
I attend to the former, using 'a # p' merely as a notation of convenience
(abbreviation) for '-(a # P)'. So we had better restate the questions.
How is what (an instance of) a # p stands for to be assayed; and what, if
anything does '#' stand for?
Turn once more to our paradigm, the fact g,(a). Its constituents
are not only the two things, g, and a, but also a subdeterminate, of the
sort I call exemplification, which (as I used to say) "ties" g, and a into the
complex that is the fact. This feature, so characteristic of my world, I
have already defended so often that I shall not here defend it again.
But consider against this background a # g,. Given the two things, a
and g,, there obviously is no need for an existent, subdeterminate or
whatever, to tie them into the complex a # g,.Or, to say the same thing
positively, given "two" things, there is eo ips0 a third, I a l l it their diad,
that is a complex whose only constituents they are, and that is actual or
potential depending on whether the "two" are diverse or the same.,'#'
thus stands literally for nothing. l 5 Such marks, of which there are in the IL
all together four, I call diacritical. They are of course the four exceptions mentioned in (Fl).
As for things, so for all determinates. There is for any "two" of
them eo ipso a third. viz., the diad whose immediate16 constituents they
are. A diad is one of the three sorts of complexes I call circumstances.
The other two we shall encounter in the next two notes. A few further
comments on sameness will fit better after these notes. First of all,
though, for two features of it.
First. Remember the condition, imposed earlier, requiring the two
terms of a sameness to be determinates. On the side of language, that
makes, say, ' v # 3'and '3# a' ill-formed. The two grounds of the
requirement are, first, on the side of the intentions, that only determinates are sufficiently "independent" for any two of them to sustain a
diad,17 and, second, on the side of the act, that only a determinate is
sufficiently "separable" to be the whole of an intention. This is of
course merely a part of the unpacking of the two metaphors (see fn. 3);
yet it should give the flavor.
Second. PM assays sameness and diversity as what the instances of a
= R L P and
stand for, even though upon this assay there will
be an instance if and only if the two determinates are things of the same
Russellian type. Hence, if one accepts Russell's assay, he cannot literally
say18 that two things of different types, or a thing and a fact, or a thing
and a class, and so on, are diverse. Yet are we not sometimes with equal
evidence presented with actual diversities of these sorts? Some may
even insist that in these cases the evidence is still stronger than in those
to which Russell limits us. T o me, who in this respect uses 'evidence' like
Brentano, the comparative makes no sense. Be that as it may, why did
so many philosophers accept Russell's assay? One cause I can think of is
the tremendous influence which, quite justifiably in many respects, he
has had on us; another is the pervasiveness of a more o r less inarticulate
nominalism.
How are instances of COUPto be assayed and what, if anything, does 'JU'
stand for? Stay with the paradigm; suppose that there is a thought, gz,
NOTES ON ONTOLOGY
141
NOTES ON ONTOLOGY
143
The search for the string or strings standing for {a, b, c) falls
naturally into two parts. We must (1) find an expression of the IL
standing for a determinate that does the job which in our natural
language is done by the enumeration clause. Call such a determinate a
selector. We must ( 2 ) state the formation rule corresponding to the
canon by which a class is built from a selector and whatever else, if
anything, is required.
(1) Pick any determinate diverse from a, b, c, say
9,(d) and consider the diad22b,(d), b,(d) = a) v
(8, (d) = b) V b5(d)= (c).Anticipate that it can do
the job and save words by calling it a selector of
{a, b, c); call 9, (d) the dummy of this selector; the
disjunction, its enumerator.
(2) Consider what has just been said about the very
nature of classes and you will see that, given a
selector, there is eo ips0 the class it selects. Yet a
selector and "its" class are two and not one. One
of the two things to be done, therefore, is to
choose a design, say 'A', make 'A' our fourth diacritical mark and write for the paradigm the string
+,
NOTES O N O N T O L O G Y
145
We are ready for what still ought to be said about the issue of diversity
and sameness.
(1) When designing an artificial notation one may
decide to introduce two marks, say again, 'a' and
'b', for one thing. In this case 'a = b' does not
stand for anything in the world but merely reflects a "linguistic rule."
(2) 'a = a' is ill-formed.
Jointly (1) and (2) paraphrase what I take to be the position the author
of the Tractatus (4.241, 4.243) has taken on the issue. Diversity and
sameness, unless I misread him, are literally nothing. If the formula is
read to assert that '=' and '.#' do not stand for anything in the world, 1
agree. Notice, though, that when it comes to 'transcribing" into the IL a
certain intention, one must look at the world and not just at "linguistic
rules"26in order to decide whether to write 'g,(a)' o r 'gl(a) . g,(b)'. That
alone suffices to convince me that the circumstances called diversitites
and therefore also their negations, called samenesses, have ontological
status. With (2), therefore, I disagree. Yet one may fairly ask of what
use, if any, are the expressions that stand in the IL for either a strict
diversity or a strict sameness. I divide my answer into three parts; the
first is general, the second and the third are specific.
First. Every ontologist using an IL cannot but be impressed
with the distinction between what can be said literally, i.e., in his
IL, on the one hand, and, on the other, although he may not put it
this way, what can be said only nonliterally in ontological discourse (see fns. 3 and 18).Nor, considering the style in which he
has chosen to philosophize, can he be insensitive to the challenge
of establishing the exact limits of what is literally sayable.
NOTES ON ONTOLOGY
147
Turning to the positive, let us first recall what the axiomatic set
theorists do about order. Starting from the class {a, b), they form the
two classes {a, {a, b)) and {b, {a, b)), call them ordered pairs, writing
for them the abbreviations '(a, b)' and '(b,a)'; proceed to define ordered
triples such as (a, (b, c)), writing for it '(a, b, c)'; and so on. Then they
prove that the classes (a, b), (a, b, c), and so on, satisfy the axioms of
(finite) linear order. For the purposes that are their exclusive concern
this will do very well; ontologically, however, this "axiomatic reconstruction" of order does not amount to an assay of it.29Its idea we owe
to Kuratowski; Bernays first fully developed it.30
This machinery of the mathematicians I take over without any
formal change. I merely shift its "interpretation," or, as I would rather
say, using a serviceable notation of convenience with Greek-letter variables, I let the latter stand for any determinate whatsoever; '(a, P)', as
once before, for the diversity or, as I also called it, the diad of a and P; '(
a. p, 7)' as an abbreviation for (a, (P, y)) and so on. As you see, I even
borrow the classical notation. But since, as I use it, (a, p), (a, P. y), and so
on, are all diads, I call them all n-tuples, in order to set them apart from
the mathematicians' ordered pairs, triples, and so on, which are all
classes. Formally, the shift seems minor, ontologically it is decisive. So I
shall next try to convince you of that and only then insert the 2-tuples
(a, b) and (b, a) in the assays of rl(a, b) and r1 (b, a), respectively.
The foundation or basis of this world are its things. From this basis
every existent is built, or generated, by a series of steps, each step being
regulated by one canon. As a result of this mode of generation, this
world is,roughlyspeaking, a layered structure. Yet, rough as the image is,
it suggests another, equally rough, such that the two illuminate each
other. Moving upward from the basis, some categories come "before"
some others; some are "closer" to the top or to the basis as some others.
Two examples will help and at the same time serve the immediate
purpose. Since, given two things. there is eo ips0 their diad, what could
be closer to the bases than certain diads? At the other end, it is very easy
to produce from the IL of our world that of one otherwise like ours but
without classes. All one has to do is to delete 'A'and 'E' and, of course. all
the expressions in which they occur. That surely puts classes at or close
to the top.
Suppose now that you have the choice between assaying the
ground of order in, say r,(a, b) by either the diad (a, (a, b)) or by the
class{{a), {a, b)).31 Upon the first alternative, which this ontology
permits, you find the building stone needed in a layer still closer to the
basis. Upon the second, you would have to reach up for it into a layer at
or close to the top. That makes the second alternative counterstructural
in a very bad sense; for it cannot but undermine the confidence in an
ontology that so "scrambles" the taxonomy of its existents. I, for one,
NOTES O N ONTOLOGY
'
NOTES ON ONTOLOGY
own, is, on the side of the IL, accounted for by such expressions as 'I=v ' 's being
ill-formed. The features of the world, or equivalently, on the side of the intentions, that
in turn account for their being ill-formed, will be more profitably discussed after we shall
have become acquainted with the four exceptions mentioned in (Fl).
SButdo not forget the familiar limitation. If the calculus contains variables, say 'x',
'y', and so on, an instance of the (half-) schema ~+(x).(x=,~y)>I+(y)
may not be analytic if a
free occurrence of 'x' in 4 lies within the scope of a quantification over 'y'.
gFor support of this judgment see later (fn. 26).
1For clinging to the ontological status of "derived characters," I did have specific
and intellectually respectable motives. Yet, having finally discovered, again at the beginningof the Seventies, how to do without them thejobs I had until then thought only they
could do, I promptly got rid of them, some time before, independently, Reinhardt
Grossmann dispatched them in his contribution to the Sellars Festschrift. ([8]): pp.
129-46).
"But notice that if in this world of 1955, say ' ra7 ' and ' rg17 'had been well-formed,
they too would have been strings, and let me use the opportunity to report still another
change of the early Seventies. In that change, every instance of aA/3became well-formed
provided only (the instances of) both a and /3 are determinates; which of course makes all
the members of several subcategories transparently potential. Yet the inconvenience, if
any, of such "ontological wasteland" is, I believe, many times outweighted by the structural, or, if you please, systematic reasons for the change.
I21n the case at hand a probable cause of this shortcoming was, I submit, that the
explication of the notion of analyticityto which I clung until the early Seventies did not, as
its successor does, contain as an essential part a painstaking phenomenological description of the conscious state to which a complex i s presented with its mode, i.e., as either analytic or
contradictoly. Nor should it come as a surprise that, on the side of the act, it involves a
notion of evidence structurally descended from Brentano's. See also the Valencia paper.
130r,in case the universal is relational or of a higher type, by either several particulars or either one or several universals of the next lower type. In this respect I am still as
good an anti-Platonist as any. But I would rather avoid the term, saying instead that, by
the Principle of Acquaintance, we cannot intend an universal unless it either comes
exemplified or has been so presented to us at an earlier occasion. For the rest, everything
not merely "speculative" (and in which I am therefore not interested) is taken care of in
the IL, by the unabbreviated equivalents, however cumbersome, of Russellian description.
14Seemy papers on Frege. ([2]: pp. 205-34 and [3]: pp. 124-57).
15Using for once a traditional phrase in order to call attention to a structural
similarity, this feature makes diversity and the other two sorts of circumstances the only
internul rehtiow of this world. Ordinarily, though, I avoid the phrase, lest it blur the
distinction between what is literally nothing, on the one hand, and, on the other, either
relational universals, which are things among things, or certain subdeterminates such as,
say, exemplification and disjunction, which, once given ontological status, may because
of their (apparent) binary feature be mistaken for relational universals. (The reason for
the bracketed 'apparent' will become clear only in the sixth note.
IeI add 'immediate' because a determinate that is a complex has itself constituents
and because in ontological discourse the use of 'constituent' is most conveniently so
regulated that constituency is transitive (and irreflexive).
"Or any other circumstance; so we shall not need to return to this feature; but see
fns. 7 and 11.
I8As I use "literally sayable" in ontological discourse, what can be literally said can be
expressed in the IL, and conversely. See also the Valencia paper.
lgThisthird comment is really but acomment on the second. But notice, finally, that
the corners, as in ' rg,(a)i ', having been introduced as "logical constants," were never
meant to stand for anything. Nor, if for a moment we resurrect that notation, do they.
That, though, does not make the notation lesscounterstructural. Rather, it shows that by
introducing two diacritical marks, 'At and ' r. . . 1', where one will do, it is redundant.
20Constituency,we remember (fn. 16),is transitive; elementhood, every one knows, is
not. Also, classes, like things but unlike complexes, have no modes. All that goes to show
NOTES ON ONTOLOGY
153
variables). But I would rather avoid 'function' in this context. Just think of the ways some
eminent ontologists have used 'function'. Yet I shall borrow "unary" and "binary."
33'=', since it occurs only in abbreviations, stands of course for nothing; yet, since it
does not occur in the IL, it is not counted as a diacritical mark.
34Fora more challenging formulation of this idea see the concluding paragraph.
3 5 Y owill
~ gather that, as already implied, for reasons of my own I ontologize not
only the syntacticists' minimum of one but all thirteen binary connectives and negation.
But, just as at the very outset we agreed that we may here take quantification without
variables for granted, so we need not pursue this piste either.