Paradoxes (Labov)

principi.qxp 09/09/2011 17.
18 Pagina 1
POLUS
RETHORICA
Serie diretta da Stefano Arduini
2
principi.qxp 09/09/2011 17.18 Pagina 2
Paradoxes
edited by
STEFANO ARDUINI
EDIZIONI DI STORIA E LETTERATURA

Prima edizione: settembre 2011

ISBN 978-88-6372-309-0
Tutti i diritti riservati
EDIZIONI DI STORIA E LETTERATURA

00165 Roma - via delle Fornaci, 24
Tel. 06.39.67.03.07 - Fax 06.39.67.12.50
e-mail: info@storiaeletteratura.it
www.storiaeletteratura.it
SUMMARY
STEFANO ARDUINI
Paradoxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
GIOVANNI BOTTIROLI
An Introduction To (Conjunctive) Scissional Logic . . . . . . . . . . . 15
PETER CARRAVETTA
No Longer a Paradox: the Sophists as Philosophers
of Language and Existence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
MICHELE PRANDI
Formal Contradiction and Consistent Thought: Oxymoron . . . . 81
STEFANO ARDUINI
PARADOXES
Giambattista Vico in his works writes that tropes are rhetorical

strategies that make one’s mind clear rather than simply tools that
provide literary effects. This statement is the crux of Vico’s analy-
sis of language. It is closely linked to another of his discoveries: all
tropes can be reduced to four: metaphor, metonymy, synecdoche,
and irony. These four universal forms or tropes build a matrix out
of which all elements of knowledge are born. Thus, metaphor is the
mother of myth, as he affirms in Notae al Diritto universale; the
development of language in childhood is the product of metonymy;
synecdoche and its special case of antonomasia produce what Vico
calls characteres heroic.
Viewed this way, figures or tropes are not deviations from standard
language but are the only true language and repose at the very place
where signs are born. Furthermore Vico affirms that figures consti-
tute not just the usual language of the poets, but also the language of
children and of primitive humans. Figures, like myths, constitute the
very language of imagination. Their home lies at the juncture where
mythical subject and object are one and the same, where the pure and
original logos can be found.
In this context Vico assigns a special role to irony, which for Vico
is the result of a conflict between two different propositions each of
which contains a possible truth. For Vico, the term ‘irony’ includes
the huge territory populated by antithetic figures and specially
paradox.
8 STEFANO ARDUINI
In his book Novantiqua Paolo Valesio1 concludes that antithesis

has a strategic importance in modern rhetoric because it permits us to
rethink the role of dialectic. In this case dialectic is not tripartite, as
in Hegel (thesis, antithesis, synthesis). Rather it is antithetic dialectic:
it doesn’t admit synthesis. In this way Valesio discounts largely what
a long Hegelo-Marxist tradition named a contradiction: a dialectic
that refers to propositions that are at one and the same time contra-
dictions and truth statements. Rejecting this view, Valesio introduces
the antithesis and affirms that this antithetic dialectic is actually a
new rhetoric.
In a similar way, Giovanni Bottiroli2 has developed the Junghian
concept of enantiodroma, that is, the antithetic way that characterizes,
for Jung (Psychologicat Types), the evolution in the life of figures like
Paul of Tarsus, Nietzsche, Raimundo Lull, and in principle other kinds
of conversion.
What in any case is important to stress is that the figures and tropes
that express a contradiction represent the world in a way that chal-
lenges what is usually accepted as true and stable. Contradiction stands
for a cognitive act and state of mind that permits us to conceptualize
the world in an unusual way. It rebels against the limits of common
language and rearranges our cognitive and semantic maps. This is an
eminently human state of affairs, as Wittgenstein wrote, since humans
feel compelled to overcome the limits of language. Contradictions,
illogical meanings, and paradoxes come from this desire.
Michel de Certeau3 notes that oximoron, like antithesis, is a sort
of creative audacity. Oxymoron breaks a code in a particular way: the
words combined in an oxymoron belong to heterogenous orders; it is a
combination of words and concepts that works differently, it is a lapsus
or a failure of simile; it blends gender and disturbs the order.
Contradiction, illogical meaning, and paradox make up the fabric
of antithesis.
1
Paolo Valesio, Novantiqua: Rethorics as a Contemporary Theory, Bloomington,
Indiana University Press 1980.
2
Giovanni Bottiroli, Retorica, Torino, Bollati-Boringhieri 1993: p. 58.
3
Michel de Certau, Fabula Mistica, Bologna, Il Mulino 1987: p. 202.
PARADOXES 9
Obviously, the subject of antithesis has many rhetorical sides: nega-

tion, inversion, irony, oxymoron, and paradox.
Paradox tests our ability to understanding reality; it produces a
creative tension which breaks certainties and permits us to see things
in a new light, with a deeper meaning.
Paradox lies at the heart of ‘Eleatic Palamedes’, which Plato con-
demned in the Phaedrus:
“Have we not heard of the Eleatic Palamedes, who has an art of
speaking by which he makes the same things appear to his hearers like
and unlike, one and many, at rest and in motion?”
As we have known since ancient times, ‘Palamedes’ is the phi-
losopher Zenon. Plato identifies the art of Zenon with the ‘antilogical’.
Although Plato condemned the ‘antilogical’ for its orientation to useless
things, he still treated it as a fundamental mechanism of his dialectic and
used it to create aporias, that is, to frame a concept or argument within a
framework or judgment that does not allow us to find a resolution.
Plato’s use of the contradiction derives from his vision of an always
moving phenomenal world which is a step towards the truth we must
discover in the phenomenal mutability of the world.
We also find antithesis in the form of paradox within Latin rheto-
ric, for example in Cicero, who dedicated his work Paradoxa stoico-
rum, to stoic paradoxes and for whom the paradox created a mental
fracture. Paradox tests our mind and creates the risk of the absurd
because it expresses human contradictions. In Cicero the paradox is
a part of Inventio and it is a technique used to bring a new argument
against what is usually accepted. In this sense it is a fight against an
established coherence that is in search of a new, superior, coherence.
Paradox is thus a fundamental process of making reality explicit: it
focuses on the contradictions surrounding us which the ordinary mind
and common opinion usually try to hide.
For example, Pascal used the figure of antithesis to describe human
beings.who are by nature subject to limits:
L’homme n’est donc qu’un sujet plein d’erreurs ineffaçables sans la grâce.
Rien ne lui montre la vérité : tout l’abuse. Les deux principes de vérité, la
raison, et les sens, outre qu’ils manquent souvent de sincérité, s’abusent réci-
proquement l’un l’autre (Pensées, 83).
10 STEFANO ARDUINI
Humans are beings who cannot distinguish between dreaming and

waking (fr. 386), swinging between will and feeling, who do not know
what is right (fr. 294), who are not able to know themselves (fr. 100), who
are changeable and superficial (fr. 366), and vain (fr. 50). A single fly can
change their thinking. ‘Oh, ridiculous hero!’ (fr. 366). However, it is due
to these limitations that humans—and here is the paradox--reach their
greatness. Humans are aware they have limits and in this awareness they
reveal their true nature, as this fragments from Pensées testify:
397. La grandeur de l’homme est grande en ce qu’il se connaît misérable. Un
arbre ne se connaît pas misérable. C’est donc être misérable que de se con-
naître misérable ; mais c’est être grand que de connaître qu’on est misérable.
398. Toutes ces misères-là mêmes prouvent sa grandeur; ce sont misères de
grand seigneur, misères d’un roi dépossédé.
409. La grandeur de l’homme est si visible qu’elle se tire même de sa misère,
car ce qui est nature aux animaux nous l’appelons misère en l’homme par où
nous reconnaissons que sa nature étant aujourd’hui pareille à celle des ani-
maux il est déchu d’une meilleure nature qui lui était propre autrefois.
In the case of Pascal, paradox turns into an anthropological tool; it
is compared to the general human condition in so far as we can only
obtain an acceptable representation of the human nature when we pile
up a sequence of paradoxes which do not find a synthetic solution. The
rhetorical figure is clearly more than a semantic rupture; the figure
defines the very nature of humans themselves: paradoxical.
A paradox tests our ability to understand things. This is why it is
one of the most favorite figures of the mystics. Paradox produces a
creative tension which breaks certainties and lights up things with a
new meaning that is not immediately visible.
Northop Frye4, reflecting on a consistent element within the mystic
tradition, writes that linguistic expressions concerning God are des-
tined to dissolve in paradox or ambiguity. In areas related to mysticism,
linguistic expressions have to include in themselves a sense of their
4
Northop Frye, Il potere delle parole. Nuovi studi su Bibbia e letteratura.
Firenze, La Nuova Italia 1994: 132
PARADOXES 11
descriptive inadequacy. Only mythical and paradoxical language can

do both at the same time, at once affirming and denying.
The rhetorical use of paradox belongs to the stylistic and theological
repertoire of many biblical texts. These biblical texts produce a crea-
tive tension that breaks definitive truths and illuminates things with a
meaning not immediately recognizable. We think for example of the
use of contradiction and antithesis by St. Paul, as in a text such as “We
seem to have nothing yet everything is ours” (2 Corinthians 6:10) (Paul
shares this love of antithesis with his contemporary Epictetus); and we
could open our Bibles to almost any page to test this claim. In a sense
the ultimate paradox of the New Testament is the Easter story: out of
death (Good Friday) comes life (Easter Sunday).
Northrop Frye, for example, sees antithesis at work as it draws a
picture of human beings in the New Testament. He remarks, “the often-
said sentence ‘Who has ears to listen, let him listen’” (Mark 4:9) is not
a message for people who have been selected beforehand for listening.
Rather it is a request that asks the reader and listener to react to Jesus
without prejudice and bias” (133). In the case of Mark 4:9 the antithesis
and the paradox transform themselves and become part of an anthro-
pological world view. They belong to the human condition. And they
show that it is possible to represent the human condition as a sequence
of paradoxes that finds no synthesis. Viewed in this way, the rhetorical
use of paradox is more than a simple semantic manipulation and cutting
apart. Rather, antithesis and contradiction become a means for defining
who we humans are: as for Pascal, paradox represent humans.
The present volume examines the topic of paradox from three per-
spectives and represents three different areas of research: linguistics,
philosophy, literature.
Michele Prandi treats a special kind of paradox: Oxymoron, which
is a figure based on contradiction. It stages a conflict that takes place
when two opposite terms are syntactically connected within a single
expression uttered by a single speaker and without temporal cleavage.
Unlike metaphor, contradiction is a formal kind of conflict, a property
that makes it suitable for framing consistent conceptual contents. A
contradiction is open to two main interpretative strategies. The first
12 STEFANO ARDUINI
dissolves the contradiction by dissociating personal responsibilities:

the opposite concepts are subscribed each by a different subject. The
second directly acts on conceptual contents: insofar as it gives simul-
taneous expression to opposite concepts, contradiction is interpreted
as a form of expression of complex, conflicting and changing states of
affairs whose description requires both the terms of an opposition.
According to Bottiroli, when we think about paradoxes we are
permited to distinguish different logical styles. The dominant style
in our tradition is the disjunctive one. Aristotle and Frege, despite all
their differences, share the same assumption: it is possible to reason
correctly only if we establish clear borders between concepts; and
the clearest borders concern opposite things; opposites are inclined to
exclude each other.
However, there is another tradition which began with Heraclitus
and drawn on by Hegel which affirms that it is absolute necessary to
recognize the connection between opposites as an ‘original link’. For
different reasons, this line of thinking, called conjunctive logic by
Bottiroli, has never brought convincing results. Among other things,
this position lacks clarity about which opposites recognize and impli-
cate each other even as they maintain a conflictual tension.
Aristotle called this kind of opposition correlative. It is not an acci-
dent that we do not find this kind of relation in the logic square, which
instead provides relations between contradictories and contraries.
Separative logic affirms that correlatives are simply ‘inconceiv-
able’ and points us to another reason to account for the difficulties and
uncertainties of the conjunctive logic.,This is the idea that a conjun-
ctive logic has to be a logic of synthesis (in Hegel’s sense). If we look
at other authors (for example, Heidegger and Freud), it is possible to
project a disjunctive logic based on connections and at the same time
on division. Obviously, for some people identity is equivalence (A = A),
but for other people identity is non-equivalence (A is not equal to A):
‘I am not what I am’, says a Shakespeare’s character, and all complex
people could describe themselves through this expression.
Peter Carravetta analyzes the use of contradiction in the sophist
tradition of antiquity. His main thesis is that we have greatly misunder-
stood the sophists owing in great part to a predominance, in Western
PARADOXES 13
culture, of Platonism and other forms of abstract idealities. The soph-

ists understood that all discourse is circumstantial, persuasive within
a given time and place, directed at actual interlocutors with specific
issues to be solved. Their understanding of human language was based
upon acceptance of a third frame of reference -- which does include
the ideas and transcendent forms -- but which was grounded upon the
circumstances of the utterance, and therefore took into account who,
what, where and for what reason an exchange was required. In this
sense, they introduced a notion of relativity which defies all axioms
and dogmas. The sophists’ view of language-in-use was a proto-prag-
matics, which belies all rhetorical theory.
GIOVANNI BOTTIROLI
AN INTRODUCTION TO (CONJUNCTIVE) SCISSIONAL LOGIC
1. Can a unified definition of paradoxes be given? Can a common

mechanism be identified and can paradoxes be traced back to a single
fallacy? There are at least two reasons for rejecting this possibility.
Firstly, it is by no means certain that a common mechanism can be
identified within the sphere of separative or disjunctive logic; secondly,
and far more importantly, because the linguistic (as well as, arguably,
the perceptive and physical) space of paradoxes must be investigated
from the standpoint of logical pluralism. By this I do not simply mean
various types of logic (such as bivalent, many-valued, deontic, tempo-
ral, fuzzy, paraconsistent etc.); all these types have a common stylistic
identity, since they make up the family of separative logic in its vari-
ety. Admittedly, their differences should not be overlooked, but they
become less marked once another possibility is taken into considera-
tion, i.e. that of types of conjunctive logic. This is a major downsizeing,
although separative differences do not collapse. Then another question
comes to the fore: is conjunctive logic really possible?
Its legitimacy was already claimed by Hegel, for example, as well as
the dialectic tradition, e implicitly also by other authors, for example,
by Heidegger. This logic was ‘put into practice’, and its working can be
perceived in some of the leading works of western philosophy. That it
is in a way perceptible or noticeable is confirmed by its very rejection:
are not the obscurity and confusion which allegedly utterly ruin the
works of Hegel or Heidegger the consequence, in the eyes of separa-
tive philosophers (such as, among others, Russell and Carnap), of their
styles of thought? As a matter of fact, conjunctive logic has never been
16 GIOVANNI BOTTIROLI
rendered transparent by any of its interpreters. What are its principles,

mechanisms and inference rules? And why is it so difficult to address
these questions? If there were no difficulty, to the extent of even induc-
ing the greatest thinkers to put off a task that can no longer be put off,
we would not be in such a weak position in the face of separative logic,
of those logical styles whose limits are clear, tiresome and unaccept-
able – for those who could never accept their unilateral working.
Rendering the working of conjunctive logic transparent is a task
that can no longer be put off: not so as to reduce or tone down the
heterogeneity between various logical-philosophical traditions, but
rather to display the legitimacy and productivity of their differences;
and to defend the novelty of ‘divided thought’, which is in danger of
being swallowed up by the obviousness of oppositions closer to day to
day mentality, and especially by fluctuations between the Unity and
the Multiplicity.
2. Initial ambiguity needs to be avoided: conjunctive logic is not

unilaterally grounded in links or bonds, i.e. forms of unity overcoming
and abolishing differences and contrasts. Conjunctive logic is scissio-
nal, not synthetic. This is my research paradigm. Therefore this essay
will foreground (starting from the title) the link between scission and
conjunction1.
1
The objection can be made that, historically, conjunctive logic has often, per-
haps mostly, been seen as a logic of synthesis. This objection may even be accepted,
but not discussed immediately, given the theoretical, not historical nature of this
essay. Nevertheless, I should like to point out that Hegel’s dialectic is less on the side
of synthesis than is generally supposed.
The bluntness of this statement should, however, be situated in the context of my
research aims. I like to think of this article as a short introduction to a text on logic,
entitled Elements of scissional logic. I chose the adjective scissional, because it seems
to me to be the most suitable for the paradoxical link between conjunction e division.
I think that the expressions logic of divisions, or divided logic, could also be used (cf.
the ‘divided subject’ of psychoanalysis). What counts is not forgetting that ‘scissional
logic’ is a paradoxical logic.
I fear that many aspects of this article may be difficult to understand without
knowledge of some of my previous work. Here I shall only mention Bottiroli, Teoria
AN INTRODUCTION TO (CONJUNCTIVE) SCISSIONAL LOGIC 17
Very important consequences immediately derive from this parado-

xical connection, albeit, for the moment, only intuitively. If ‘conjunctive
divisions’ exist, then so do ‘separative conjunctions’. Therefore, the
meaning of terms such as distinction, division and conjunction can only
be understood with reference to a struggle between styles of thought.
Let us go into this line of thought in greater detail.
All styles introduce articulations. Each style articulates the flux of
language-thought differently2. To articulate means ‘to distinguish’: and
since there are different types of articulation, a distinction can be separa-
tive or conjunctive. Separative distinctions enclose individuals and objects
within their borders, even when they point to relationships of solidarity,
strong resemblance or reciprocity. For example, the same makes and
types of car (with the same optional extras and painted the same colour,
etc.) can still be easily distinguished; Luca’s identity is not diminished
by the fact that he is Gabriele’s brother, even though ‘being someone’s
brother’ is a relationship, not a property; and so on. In the field of separa-
tive distinctions, numerical identity always allows counting and separat-
ing, i.e. avoiding confusion; and when confusion comes about, regarding
a person or object, it can always be entirely dispelled3. It does not appear
to be an overstatement to say that all distinctions concerning identity,
in the separative regime, rotate around numerical identity. Conjunctive
distinctions, on the other hand, are grounded in reciprocity in a different
way: not that numerical identity is absent, though its importance is drasti-
cally reduced. The notion of ‘identity’ thus needs to be redefined in an
intrinsically relational way: conjunctive identity is flexible.
dello stile, 1997, and “Metaphors and modal mixtures”, in Arduini 2007, pp. 17-41.
This essay can be consulted at www.giovannibottiroli.it.
2
For a view of language grounded in articulations (actually not differentiated
by different types) cf. the double-flux scheme in Ferdinand de Saussure, Cours de
linguistique générale, 1916.
3
Vague borders are not incompatibile with rigidity. There can be considerable
doubt about a particular piece of land belonging to the Matterhorn, but this appears to
be a problem of interest only to some philosophers of language: where the Matterhorn
is situated, where it should be looked for – either in nature itself or on a map – leaves
nothing at all to chance.
In the conjunctive regime, the opposition between rigidity and flex-

ibility is overwhelming. But let us not be too hasty: these preliminary
clarifications appeared necessary so as to open up to intuition of the
complexity of the problems involved; let us now start off from the ‘con-
junction’ phenomena appearing in paradoxes.
3. Many scholars see paradoxes as providing the best confirmation

of the unavoidability and superiority of separative thought. If the reas-
surances that only this kind of thought can provide disappear, then nulla
salus. We are entering uncertain territory, where every project and inten-
tion of coherence will inevitably be abandoned. Nevertheless, nothing
forces us to enter or lose our way in excessively intricate places. There
will be the chance of building up an area which cannot be encroached on
by logical errors, and from which any possible error can sooner or later be
driven out. The reader will easily recognise the reform of natural language
begun by Frege in this description. As is well known, the aim of this pro-
gramme, i.e. paraphrase of natural language in a language clarifying and
showing its true form, developed in two contrasting, though essentially
parallel versions. Some see paraphrase as a reformulation which does not
completely discredit ordinary expression, while others see the need for
systematic correction4. Nevertheless, a separative philosopher can easily
be recognised by his attitude to paradoxes: errors, sophisms, knots to be
untied, which once done cannot be tied up or come to the fore again
This attitude is coherent with the rigidity of this style of thought: any
violation of borders, which are ‘good separations’ is a lapse into the con-
fusive – is this not the name we should use for conjunctive thought? On
the other hand, those who do not identify themselves with the separative
conception, those who mistrust it or are irritated by it, will be tempted
to begin the search for unresolvable puzzles. They may well exist and be
pointed to as a limitation on separative hybris. Actually, the discovery
of unresolvable paradoxes would be a poor result (just like all defensive,
or in Nietzsche’s words, reactive outcomes). What is really important is
4
Cf. for example Varzi 2001; however, the terminology used (interpretation and
revolution) is unsatisfactory.
discovering that, alongside ‘chained’ paradoxes, there are others whose

legitimacy and productivity should be acknowledged. Is there, or could
there be, a way of thinking where chains are not chains, where they are not
constraints that slow down or act as stumbling blocks, but rather are the
support and precondition for more agile movements, excluded from separ-
ative thought? “Dancing is a mistake, let us restrict ourselves to walking”:
is not this, in the end, the cautious fallacy of disjunctive thought?
4. We appear to have taken on a ‘symtomal interpretation’ not only

of paradoxes themselves but also of the ways of addressing them. What
do paradoxes display or reveal syntomally? Is it the imperfection of
ordinary language? Is it confusion, aporias, pathologies, from which
one can defend oneself by building up a language capable of never
being deceived over its own logical form? Or a set of links that can
only in some cases be successfully investigated by separative logic?
But to strengthen this hypothesis we will have to increase the list of
paradigmatic cases and find genuinely ‘conjunctive’ ones: i.e. situa-
tions where the intervention of separative procedures turns out to be
not impossible – unfortunately, in a flexible context like that of lan-
guage, bad rigidity is always possible – but rather sterile, short-sighted,
even stupid.
What should therefore be rejected is preliminary domestication of
the problem. But where should we search for paradigmatic cases capa-
ble of making the failure of a specific type of logical paraphrase clear?
For example, in a type of language undervalued by Frege’s research
paradigm, relegated to the emotional sphere, a forgotten language,
even as an object of criticism and correction.
In the words of Wittgenstein: “What we do is to bring words back
from their metaphysical to their everyday use”5. That here we have the
rejection of a search for an artificial language, disastrously impover-
ished by its own hygienic ideal, is, without doubt, of enormous impor-
tance, but it is not enough. It is Wittgenstein’s alternative that seems
5
“Wir führen die Wörter von ihrer metaphysischen, wieder auf ihre alltägliche
Verwendung zurück” (Wittgenstein 1953, Part One, 116, p. 41).
unacceptable. In opposition to this, one could well provide the follow-

ing parody: “We aim to bring words back, from their metaphysical or
everyday use to the aesthetic dimension. We want to look upon and
judge all language from the viewpoint of the languages of art”. From
Wittgenstein, we want to return to Nietzsche.
5. What I shall now present is a conceptual research programme,

rooted in types of philosophy, which, as has already been said, have
never tried to make their logical working transparent. I look forward
to receiving constructive suggestions, so as to add to and improve
a proposal that I do not see as strictly individual, but rather open to
cooperation. As mentioned at the outset of this essay, investigation
of conjunctive logic can no longer be put off: it would be particularly
beneficial to all those who, from different viewpoints, have in common
their rejection of the separative style.
Looking forward to ‘technical’ cooperation does not mean aim-
ing at formalisation: that conjunctive logic can be reformulated in the
language of symbolic logic is very doubtful, to say the least. One has
the feeling that a flexible logic cannot be formulated by means of a
language primarily grounded in rigidity. Partial transposition, limited
to only some aspects, could, however, be plausible; though this should
not be done in a hurry. Placing technical aspects before conceptual
analysis would be a very serious mistake.
Besides, is not conjunctive logic called upon to overturn Frege’s
paradigm? Should not one of its aims be to show the fallacies through-
out symbolic logic? It should be noted that a fallacy is not an error.
The two notions can be confused, as often happens among analytical
philosophers, since they restrict themselves to ‘local’ fallacies6. But
the word fallacy ought to be used predominantly to cover deformations
or distortions, which cannot be limited to mere, locally identifiable
sophisms. This word ought to be used to mean widespread errors,
6
For example petitio principii. Or ignoratio elenchi: “This crime which Bill is
charged with is horrible; so Bill should be found guilty”. For this and other examples,
cf. Penco 2005, p. 15.
so widespread as to be difficult to perceive – unless one takes a step

backwards to return to logical pluralism, the conflict between styles. I
think that the greatest fallacy in analytical philosophy is monostylism,
the superstition of separativism. But analytical philosophers are not at
all aware of this superstitious faith.
Let us now return to the need for expanding the field of investiga-
tion, and finding paradigmatic cases other than those of the Liar or
Sorites or the class of all the classes that do not belong to themselves.
We have decided to look for conjunctive (not merely confusive) para-
doxes in literature, but the working of flexible logic should also be
investigated in the field of explicitly strategic relationships, i.e. in
the art of warfare and politics, and, to some extent, psychoanalysis. I
should also like to admit to being curious about physics.
This curiosity is not, however, that of a specialist. It may well be
that no conjunctive phenomena can be found in the research domain of
physics or other natural sciences. I would be disappointed if this were
so, but this would clearly not lead me to modify the theses advanced
so far; their legitimacy is to be searched for in the logical pluralism of
the human mind and in flexible languages. It would, however, be very
interesting to receive an affirmative answer to the question: “Are para-
doxes present in the world of physics?” from specialists. The reasons
for this interest should be clear. There are paradoxical statements to
which nothing corresponds, as in the cases of contradictions. There is
nothing that can correspond to “James is sitting and at the same time is
not sitting”, and the same can be said concerning the barber paradox7.
The non existence of a reference in the effectual dimension appears to
confirm the opinions of those who consider paradoxes playful, or only
ephemerally serious: they will be resolved sooner or later. The obstacle
will disappear into the void whence it came.
The most stimulating paradoxes do not allow themselves to be
restricted to the linguistic dimension but rather concern reality, in a
broad sense, and these include the ones concerning personal identity.
This is why physics can also be of interest to literary scholars, and
7
Odifreddi 2001, p. 147.
philosophers who find logical inspiration in literature, and, more gen-

erally, the arts. If there are paradoxes in the world of nature, physical
reality is working paradoxically. We will have further confirmation of
the existence of paradoxes that do not originate in the bumps of the
mind when running up against the limits of language8. Could conjunc-
tive logic be relevant to the description of certain physical phenom-
ena, or even the whole of nature? Has the unbroken wholeness of the
universe anything to do with Nietzsche’s idea that “all is enlinked,
enlaced, enamoured”9?
It is the specialists’ task to clear up the matter of whether we are
dealing with true affinities or merely fascinating, though ambiguous
resemblances. For example, should Einstein’s principle of separability
be included in the domain of separative thought, in accordance with the
definition given in this essay? And what kind of logic could adequately
describe the unbreakable systems of quantum mechanics10?
6. Let us take a quick glance at the paradoxes discussed especially

by logicians and philosophers of language. They are very numerous:
some scholars appear to be resigned to their great variety, while oth-
ers have opted for listing them selectively in alphabetical order11. The
number and variety of paradoxes, especially when presented without
taking into account their degree of complexity and mixing the more
serious cases with the more playful ones, give rise to a feeling of ran-
domness; and yet, if one highlights the better known cases, some kind
of family resemblance does emerge. It is in the latter feeling that ele-
8
Reference is clearly made to Wittgenstein: “The results of philosophy are the
uncovering of one or another piece of plain nonsense and bumps that the understand-
ing has got by running its head up against the limits of language. These bumps
make us see the value of the discovery” (Die Ergebnisse der Philosophie sind die
Entdeckung irgend eines schlichten Unsinns und Beulen, die sich der Verstand beim
Anrennen an die Grenze der Sprache geholt hat. Sie, die Beulen, lassen uns den Wert
jener Entdeckung erkennen) (Wittgenstein 1953, Part One, 119, pp. 41-42).
9
“Alle Dinge sind verkettet, verfädelt, verliebt” (Nietzsche 1883-1885, 4th Part,
“The Drunken song”, p. 313).
10
Cf. Klein 1991, pp. 162-163.
11
Cf. Odifreddi 2001 and Clark 2002.
mentary typologies originated, such as Ramsey’s division between set-

theoretical paradoxes (e.g. Russell’s paradox) and semantic paradoxes
(e.g. the Liar). As is well known, Bertrand Russell was not convinced of
the usefulness of this division and held that all paradoxes arose as the
result of one fallacy, from violations of the ‘vicious circle principle’12.
This is probably not the case: various mechanisms should be identified:
reference to infinity, self-referentiality and links between opposites.
Not all paradoxes originate in self-reference and there are many harm-
less self-referential sentences, such as “This sentence is in English”.
Besides, if we consider the fact that some mathematical arguments,
including Gödel’s demonstration of the incompleteness of arithmetic,
make use of self-referential sentences, the damage that would be done
by the elimination of reflective mechanisms is quite clear13. As far as
the link with infinity is concerned, this only characterises a certain
number of paradoxes: the most conducive to arithmetical solutions.
We still have the link between opposites, in various shapes and
sizes: the circular structure, as in the case of the Liar paradox or that
of the postcard; Orestes’ dilemma, who was destined to commit a ter-
rible crime, i.e. murder his own mother to carry out an act of justice,
revenge his father’s murder etc. We should not really be influenced by
the family air in these examples but concentrate on the differences.
New distinctions should be tried out: for example, why not accen-
tuate the difference between the paradoxes of finiteness and infinity?
Zeno should not make us forget Heraclitus and the paradoxes pregnant
with the tragic spirit and conflict: but what conflict could there be in
infinity? Does not being attracted by infinity depend on a desire to
suspend conflicts and struggle? An arrow flying through the air and
staying still, Achilles who never catches up with the tortoise. One is
tempted to say: let us leave infinity to the theologians and mathemati-
cians and occupy ourselves with conflictual relations and strategy. Let
us devote our energy to clarifying the paradoxes of flexibility, which
elate and do not create ties, or rather do not chain up. And why not
12
Cf. Haack 1978, p. 138.
13
Ibidem, p. 139.
thoroughly explore the hypothesis that chained paradoxes, i.e. aporias,

all originate in rigidity?
7. It is not difficult to find immediate confirmation: for example, in

order to resolve the sorites paradox fuzzy logic is enough. But fuzzy
logic is not the logic of flexibility, which is the object and aim of our
research.. Nonetheless, fuzzy logic is a somewhat more flexible kind
of logic than those that have dominated separative thought up to the
present14. Let us look at another example, which is conceptually more
weighty: the problem of how to learn a rule (understand and apply it).
Since any rule can be interpreted and applied incorrectly, there appears
to be the need for a metarule to direct the mind correctly. But to assimi-
late a metarule correctly, another metarule will be required, and so on.
Does not this movement backwards ad infinitum, discussed by Kant and
Wittgenstein, originate in rigidity? Only a rigid form of understanding
feels the need to learn, as Kant observed, what no school can teach15.
And this would be a source of paralysis. The solution of the paradox
should not be sought by insistence on distinction and multiplication of
levels – this can give rise to fallacies which separative rationality is
unable to grasp – but rather in pluralism of modes (or styles) of ration-
ality. No metarule could prevent a rigid subject from falling into the
net invented by the suppleness of the enemy.
Can this conclusion be generalised? Perhaps. But it must be pre-
cisely expressed, nevertheless. Once the hypothesis that all chained
paradoxes originate in rigidity has been posited, we are still at the
initial stage: rigidity displays its effectiveness in many ways, and we
must be able to understand it in detail. Let us therefore concentrate on
14
Cf. Kosko 1993. On the sorites paradox see pp. 93-97. In more ways than one,
fuzzy logic appears to be the caricature of an authentic flexible logic. Admittedly, it
does not allow itself to be inhibited by traditional rigidity, but the simplistic route it fol-
lows is still that of the intermediate case. As to truth, it is said to have a grey nature (a
mixture of white and black, p. 101). What about paradoxes? They are half truths: “The
paradoxes are literally half true and half false” (p. 80). This is not the case: the simplistic
stance of fuzzy logic will become clear when we analyse the typology of opposites.
15
I. Kant, Kritik der reinen Vernunft, 1781-1787.
a single paradox. If we manage to clarify its mechanism better than

our rivals – a long tradition dominated by separative thought – , our
analysis will immediately take on a paradigmatic value.
We will examine the Liar paradox, which recalls extremely impor-
tant notions such as ‘truth’ and ‘meaning’ and initially discuss the best
known solution, that by Tarski. We will consider the most significant
aspects:
a) the semantic concept of truth. Tarski posits factual semantics ori-
ented in the direction of actuality: “Semantics is a discipline which,
speaking loosely, deals with certain relations between expressions of a
language and the objects (or “states of affairs”) “referred to” by those
expressions”16. Tarski adds:
As typical examples of semantic concepts we may mention the concepts
of designation, satisfaction, and definition, as these occur in the following
examples:
the expression “the father of his country” designates (denotes) George
Washington;
snow satisfies the sentential function (the condition) “x is white”;
the equation “2.x = 1” defines (uniquely determines) the number ½” .
b) the property conception (let us call it that) of truth. This means that
‘true’ is considered a property characterising certain sentences17;
c) the relationship with the conception of truth as corresponding to
facts. This is a controversial point. In Popper’s view, Tarski had reha-
bilitated the theory of correspondence18: given any recurrence of the
T-schema, e.g.
“snow is white” is true if, and only if, snow is white
Popper appears to argue that the left side refers to language and the
right to facts. This interpretation has often been criticised, since it is in
16
Tarski 1952, p. 17.
17
the word “true” ... expresses a property (or denotes a class) of certain expres-
sions, viz., of sentences (ibidem).
18
Popper 1960, pp. 223-224.
contrast with what Tarski writes and the originality of his presentation.
In the T-schema
X is true if, and only if, p
the symbol ‘p’ stands for an arbitrary sentence of our object-language,
while the symbol ‘X’ represents the name of the sentence19. Should we
therefore consider Popper’s interpretation a total misunderstanding?
The question is complicated by Tarski’s ambiguity. On the one hand,
he states that the T-schema is epistemologically neutral, i.e. in respect
of the debate between realism and idealism, while, on the other, he
declares that “our formulation does conform to the intuitive content
of that of Aristotle” and also conforms to common usage, to the point
of view of everyday life20. So, if the originality of Tarski’s argument
consists of shifting the problem of truth from the relationship between
language and reality to an (equivalence) relationship within language,
these statements do not support this originality.
So Popper’s interpretation does not appear to be a banal misunder-
standing, but rather an interpretative decision. Arguably he understood
the implicit solidarity between Tarski’s theory and the traditional
theory of truth as correspondence; and his decision appears the more
plausible if we examine the T-schema in any saturated version, where,
for example, «Snow is white» can be understood as a sentence and not
a noun.
Tarski’s argument can be more easily understood if we examine
the T-schema in an unsaturated version (X is true if and only if p). We
can take a close look at Tarski’s aim, i.e. define the notion of ‘truth’ by
means of that of ‘satisfaction’21.
d) what are the advantages of this conception? It must be acknowledged
that the T-schema is impregnable, and, in any case, more perspicuous
than the notion “in agreement with reality”. In support of his approach
Tarski also mentions a test:
19
Tarski 1952, p. 22.
20
Ibidem, p. 32.
21
Ibidem, pp. 24-25.
I was by no means surprised to learn that in a group of people who were

questioned only 15% agreed that ‘true’ means for them ‘agreeing with reali-
ty’, while 90% agreed that a sentence such as «it is snowing» is true if, and
only if, it is snowing22.
Naturally this test is not of crucial importance and one cannot fail
to observe how Tarski’s ambiguity comes again to the fore. Here he
is undeniably referring to the relationship between a sentence and a
fact.
e) the distinction between object-language and metalanguage.
According to Tarski, this distinction allows one to avoid or resolve
paradoxes, and, in particular, successfully address the liar paradox. So
Tarski relies on language hierarchy: the truth of a sentence for a given
level is always expressed by a predicate of the subsequent level (if O
is the object-language, metalanguage M will contain the means for
reference to the expressions of O, and thus the predicates “true-in-O”
and “false-in-O”, and so on). The liar aporia “This sentence is false”
will then take on a harmless form “This sentence is false-in-O”; the
latter sentence will belong to M, and hence cannot be true-in-O, and is
simply false instead of paradoxical23.
8. Let us admit, for the moment, that Tarski’s conception is error

free. It can be blamed for a much more serious flaw: it condenses a
series of fallacies and distortions – it resembles someone wearing such
tight shoes that he can only walk with difficulty, in a stiff and awkward
way24. All in all these fallacies are a perspective error in the working
of our intelligence. I shall now attempt to justify such a serious accusa-
tion. The main distortions in Tarski’s conception are:
i) the semantics he posits is factual, i.e. heteronomous, only plausible
in a modal mixture dominated by actuality. Tarski denies meaning
any chance of freeing itself from a reflecting, servile elaboration; he
22
Ibidem, p. 32.
23
Cf. Haack 1978, pp. 143-144.
24
The metaphor comes from Wittgenstein 1977.
denies the autonomy of meaning. This fallacy is opposed by what I

shall call Nietzsche’s law, i.e. the non-derivability of the semantic from
the factual: this is how I suggest the well-known thesis “there are no
facts, only interpretations” should be understood25. Non-derivability
does not exclude partial derivability: it does, however, affirm the
autonomy of the semantic, the possibility for meanings to find sup-
port in the empirical to take on unexpected identities, not modelled on
sentences such as “Snow is white” (one could use the term anaclisis
principle (Anlehnung), thinking of how, in Freud, sexual drives depend
on self preservative ones so as to become autonomous thanks to greater
flexibility26;
ii) the T-schema is probably watertight but vacuous. It is tautological,
has nothing to say to us about truth, either from the philosophical or
epistemic viewpoints27. One could answer the charge of philosophical
emptiness by observing that Tarski did not intend to make a philo-
sophical contribution, and had stated that he could not understand what
‘the philosophical problem of truth’ was28. But this reaction presup-
poses an inadequate, stereotyped concept of ‘philosophy’, in the sense
of the search for generalities and essences, and a discipline incapable
of entering the technical domain. Undoubtedly, philosophy should not
be limited to technical discussions; if this were the case it could only
criticise errors and not fallacies. The expression ‘the philosophical
problem of truth’ already seems less mysterious if use is made of this
distinction. What the reader is reading is a philosophical essay because
it aims at foregrounding, among other things, a number of fallacies
concerning truth;
iii) there are areas of research where philosophy takes account of
technical details, albeit in a wider perspective. Tarski’s trust in the dis-
tinction between object language and metalanguage was undermined
F. Nietzsche, Nachgelassene Fragmente 1885-1887.

25
Freud 1915-1917, pp. 351-352. However the idea was already present in Three
26
Essays on the Theory of Sexuality, 1905.

27
Cf. Engel 1998.
28
Tarski 1952, p. 33.
by Kripke29. Let us consider these two utterances, made by Jones and

Nixon respectively:
(1) “All of Nixon’s utterances about Watergate are false”
(2) “All of Jones’ utterances about Watergate are false”
An appropriate metalanguage level cannot be identified in this case:
(1) is a higher level than Nixon’s statements, but (2) is at a higher level
than the statements by Jones. We must acknowledge infinite pursuit as
in the postcard paradox.
Have we again come up against a technical difficulty, from which
we can only escape with an adequate technical artifice? Or is wider
ranging philosophical reflection required? What is the limit we should
become aware of, if we want to go beyond it? The discussion we are
referring to – other names, such as Russell, could be added here to those
of Tarski and Kripke – is entirely restricted to domain of separative
thought, and trust in specific distinctions: so as to exclude all superim-
positions, whatever the dimension, horizontal or vertical, in which they
are articulated. The hierarchical construction of levels is one of the most
characteristic options of the separative style: when a difficulty comes
up, it is attributed to a loss of the correct distance, a shortening, or, as
in the example of the Liar and that of Kripke, to a gap or interruption in
the hierarchy. The aim of the separative logician does not change: i.e. to
find a separation point, which, in turn, should interrupt the vortex.
There is never suspicion of a levelling fallacy, i.e. a disease caused
by the cure. And yet the search for a meta level can lead to patholo-
gies: is this not the case – as we have seen – in the retreat ad infinitum
deriving from the pursuit of a metarule? It is rigidity, and not infinity
that causes paralysis. This awareness gives rise, in some scholars, for
example Lacan, to rejection of metalanguage. This does not mean that
internal distances can never be created in language, that language can
never be spoken about through language: the levelling fallacy, which
becomes a metalinguistic fallacy, depends on the fact that there is no
object language like the one described by the separative logicians,
except in the domain of formalised languages. To be more precise:
29
Kripke 1975.
no language can become an object language (in the separative sense)

which can be thought as a divided language30.
iv) every fallacy is an epistemological obstacle which blocks research,
at the same time simplifying and weakening it. The levelling fallacy
prevents access to the new space of divided language and its initial
exploration. The property fallacy – because the property conception,
which has already been mentioned, is also fallacious – consists in treat-
ing the ‘true’ predicate as an empirical predicate. However, it is only
a misleading grammatical resemblance that can lead one to believe
in a predicative analogy between “snow is white” and the sentence
“«snow is white» is true”. This consideration is undoubtedly accepted
by other scholars. Nevertheless, the strangeness of the ‘true’ predicate
appears to have been mostly perceived as a symptom of its superfluity.
In the deflationists’ view, the redundancy of the word ‘true’ should be
acknowledged on the basis of the fact that it makes no essential contri-
bution to thought with its meaning:
If I assert “it is true that seawater is salty», I assert the same thing as if I assert
that «Sea water is salty”31.
This quotation from Frege, who cannot be seen as a deflationist,
could recall a remark by Kant concerning the ‘emptiness’ of the modal
categories32. They do not contribute to the semantic determination of
entities; but can we neglect the difference between an object thought
as only possible and one thought as existing? Certainly not. The fact
that the ‘true’ predicate does not contribute to the determination of
meaning does not justify its elimination; but what contribution does it
make? Must we look for it at the level of force, and see it as an assertion
operator? Motivation seems plausible but also weak: the ‘philosophical
problem of truth’ is trivialised, entering an area of everyday tedious-
ness; the fact that the ‘true’ predicate cannot be suppressed ends up
The stylistic perspective is decisive here: style means divided language (cf.
30
Bottiroli 1997).
31
Frege 1915, p. 252.
32
I. Kant, Kritik der reinen Vernunft, 1781-1787.
by taking on routine features, resembling the need to devote time to

shaving every day or at least every now and then.
Let us take a second look at the hypothesis that ‘true’ is a modal
predicate. What is the cognitive role, the cognitive contribution of its
emptiness? It should not be looked for on the content (things, proper-
ties, facts, events) level, since true has a perspective meaning. If it
cannot be eliminated, this is because it is uneliminable like perspec-
tives, modes and styles – which only zero stylistic (zero perspective,
i.e. separative) thought can be under the illusion of eliminating.
To justify these statements one can obviously appeal to Nietzsche
and Heidegger, to truth as aletheia rather than adaequatio. But it does
not appear to be easy to persuade an analytical philosopher to spend
some time reading Heidegger, so a swift, concise argument should be
found. The need for a modal viewpoint on the problem of truth could
be illustrated by an example aiding reflection on the limits of symbolic
logic as a form of thought. As is well known, one of the most inno-
vative aspects of Frege’s conception consists in doing away with the
subject/predicate pair in favour of another conceptual pair: predicate
and argument. Every proposition is primarily seen as an unsaturated
formula of the type “x conquered Gaul”; this formula, or propositional
function, will become true if we replace x with Julius Caesar, and false,
if we replace x, for example, with Alexander the Great.
No doubt. But let us look at another example:
On one occasion Sherlock Holmes asked me: “My dear Watson, what do you
deduce from the fact that you can see the starry sky above you?”. I answered:
“I deduce that my position faces north; actually constellations in a cloudless
sky allow one to take one’s bearings not only at sea, but also in the moun-
tains, where we happen to be”. Satisfied with my answer, which sounded at
the same time profound and complex, I smiled at my friend and companion
in so many misadventures, who answered, calmly lighting his pipe: “My dear
Watson, you never cease to astonish me; didn’t you notice that our tent’s been
stolen?”. By Jove, I never thought of that!33
33
I quote this joke from the introduction by Carlo Penco to the volume he edited,
La svolta contestuale, 2002, p. XIII.
Why does Watson’s answer sound funny? Not because it is false, or

even inexact, but because it is not relevant. He answers the question,
which, when formulated in Frege’s terms, would be “what value of x
satisfies the propositional function «I deduce x when seeing the starry
sky above me»?”, “that my position faces north etc.”. Overlooking the
fact that being able to look straight up at the sky should suggest that he
and his famous companion had been the victims of theft, Watson says
nothing false. His answer, from the viewpoint of truth-values, correctly
saturates the above mentioned unsaturated question. Correctly, but also
foolishly.
What Frege’s conception neglects – and throughout the 20th century
the separative logicians were never aware of this gap – is the difference
between saturation modes. Frege does not distinguish between what is
‘foolishly true’ and truth accompanied (or modalised) by intelligence.
He does not consider cases such as “x foolishly satisfies p”.
Could we make up for this gap by means of fresh attention to con-
texts? Taking account of context, linking exactitude and relevance, is
certainly an important step forward, but it is entirely unsatisfactory, if
contextual differences remain within separative rationality, and if there
is no progress from differences between contexts to style division.
9. It is worth pausing briefly over the limits of the pragmatic turn

observable in linguistics as well as logic, over the last few decades.
It was inevitable that, initially, focus should be directed on relations
and structures within language: thus the principle of context, stated
by Frege, turned out to be somewhat less contextual than it could
have been, and not sensitive enough to extralinguistic contexts. It
would, however, be rather trivial to believe that the pragmatic turn
had coincided with the discovery of flexibility as autonomous ration-
ality, and with the acknowledgment of logical pluralism, and that it
aimed at going beyond the intention of protecting rigid reason from its
awkwardness. So there was no revolution, just an inevitable, cautious
evolution, overseen by common sense.
It would also be trivial to believe that the saturation mode problem
could be addressed, apart from being introduced, only by means of
cases with humorous effects; and that the modal dimension of thought,
which symbolic logic is unable to analyse, even when dealing with

modality, should be exclusively identified with the problem of satura-
tion modes. Given the introductory nature of this essay, it is wise not
to range over too wide a field and return to paradoxes as a symptom of
‘relational’, and not only aporetic thought.
We had decided to concentrate on the liar paradox; from there atten-
tion has shifted to a piece by Tarski, where numerous fallacies were
identified, among them the levelling fallacy. The distinction between
object language and metalanguage seemed valid in restricted environ-
ments: it is not, however, effective in cases of circularity like the one
pointed out by Kripke, which has some affinity with the liar, but could
still be put to use in the latter case. Why is this? Because the separa-
tive can undeniably be successful when in contrast with the confusive:
but the confusive is only a possibility of conjunctive thought. So other
paradigmatic cases must be sought.
Other cases of correlation: in the liar paradox the reciprocal
implication of true and false can be interpreted as a specific kind of
relationship between opposites. And could not the correlatives be the
opposites which presuppose each other, as far as existence and defini-
tion are concerned? This is how Aristotle posits them in a typology that
deserves further investigation after a long period of neglect. But let us
move forward in an orderly fashion.
10. Let us try to see the Liar paradox as a node between opposite
terms, implying instead of excluding each other, and not as a confu-
sion of levels determined by self-reference. How can this come about?
The first thing we need is a map allowing us to find our way through
the intricate polysemy of opposite relationships: an area partially, but
constantly submerged by the confusive, by intricacies, whose fascina-
tion it is difficult to avoid, and which one tradition calls coincidentia
oppositorum. The difficulty of analysing these intricate areas of thought
is undeniable, and one can understand the impatience with which the
separative mind felt both the desire and need to keep them at a distance,
with decisive strokes of an axe cutting through tangled knots and mark-
ing out an exit. We can find the first, exceptional description of this
‘dark forest’ in Aristotle and his table of opposites (B>OUJLFJsNFOB):
– contradictories (B>OUJsGSBTJW)
– contraries (UBrF>OBOUJsB)
– privatives/positives (TUFsSITJWLBtFYJW)
– correlatives (UBrQSPsWUJ)34.
The contradictory relationship concerns incompatible opposites:
“Socrates is sitting” and “Socrates is not sitting” are propositions
excluding each other, since they refer to the same individual at the
same time. The contrary relationship is weaker and permits interme-
diate cases: white and black can mix and produce grey. The privative/
positive relationship is exemplified by “not having/having sight”. As
for correlatives, the examples given by Aristotle are “half/whole” and
“master/slave”; the characteristic of correlatives is reciprocal presup-
position: one can only think of a master in a relationship with a slave,
and vice-versa. So the relationship between correlatives immediately
comes to the fore in its paradoxicality: here opposites are constrained,
though not actually enchained, by each other; by opposing each other
they necessarily recall each other.
In this brief presentation, taken from Aristotle, ambiguity can easily
be glimpsed: the examples used cover both isolated notions (“white”,
“half”) and propositions (“Socrates is sitting”). The difference is any-
thing but negligible, since one isolated notion cannot be judged either true
or false; the minimum dimension of truth consists of a proposition, i.e.
the link between two notions (or concepts). This ambiguity is suppressed
in the construction in which medieval logicians returned to the four fun-
damental types of assertory propositions identified by Aristotle: in the
square of opposition there are only propositions and not single terms.
In this arrangement, however, the fullness of Aristotle’s typology is
diminished. The ‘privative/positive’ relationship is no longer present,
and this is not too serious, because it is the least interesting one from
the conceptual point of view: Aristotle assigns specific limitations to
it, connecting it with natural conditions35 and characterising it as a
34
Aristotle, Categories, 11b 18-19 and Metaphysics, V, 1018 a 20-21.
35
Privatives are the negation of what should be a natural condition (“Not what
is without teeth or sight do we, therefore, call toothless or blind”, Categories 10, 12
a 31-33).
univocal shift from positive to privative and not vice-versa. But the
correlative relationship is also absent. The contradictory and contrary
relationships are left and their affinity is to be identified in the fact that
they both foreground reciprocal exclusion. In the relationship between
propositions there is no room for compromise, as is the case, on the
other hand, when considering isolated terms: the contrary relationship
between notions permits intermediate cases; a possibility that appears
to be somewhat different from the interdependence of correlatives,
because white is not forced to mix with black. Their reciprocity (as
parts of the same paradigm, in the sense of the term as used by lin-
guists) is differential, and not conflictual or strategic. Admittedly the
relationship between half and whole is not conflictual either, but this
means that Aristotle’s typology needs to be improved, not put aside.
The medieval logical square is, stylistically, a separative construc-
tion: each constituent, placed in one of the four corners, is clearly sepa-
rate from the others; opposites are well separated, so as not to create
paradoxes or aporias. Paradoxes are always aporetic, from the separa-
tive point of view: knots to be untied, confusion to be eliminated. This
is why the case of correlatives was eliminated: in this relationship
opposites are no longer separated or separable. Quite the reverse: they
are inseparable, defined by their conjunction.
It should be noted that close proximity and mixing between oppo-
sites does not in itself involve any particular risks: for mixtures to be a
clearly derived and not originary case, thought will admit and observe
them with no worries at all. But when opposites affirm their reciprocal
dependence, when their tie appears decisive for defining each one’s
identity, then separative logic perceives a risk of confusion and dis-
solution, even in its own case. Being powerless to analyse this type
of relationship, it feels overwhelmed by it. The reaction that follows is
entirely consequent: correlatives, the ‘non separatives’ must be driven
out into the domain of the illogical.
The removal of the correlatives, ratified in the medieval square
diagram, was to be inherited by the dominant tradition in western phi-
losophy: for example, Kant would allow only two kinds of oppositive
relationship: logical contradiction and real opposition (Realrepugnanz),
which reflect Aristotle’s contradictory and contrary relationships.
From German idealism onwards, however, the removal bar was raised;
different working of two modes of intelligence was foregrounded:
understanding (Verstand), the faculty operating through disjunctions
and reason (Vernunft), the faculty operating conjunctively. The sterile
nature of the ‘A = A’ relationship was criticised and a step forward in
respect of the principle of non-contradiction theorised, thus confirming
the (implicit?) fears of the separative philosophers: the necessary inter-
dependence between concepts, the relationship between correlatives,
threatens the supreme logical principle. A real threat, on the one hand,
though unreal on the other: real and dangerous like all infection, unreal
and unfounded like a delirium. The attacks of dialectic thought on the
principle of non-contradiction are merely fanciful and senseless.
Actually, if conjunctive logic were a “logic of contradiction”, if
it really managed to suspend the principle defined by Aristotle for
the first time, how could it be justified? In the name of a mimesis in
respect of the contradictions allegedly making up reality, especially
social reality? But social conflicts do not exemplify contradictions in
the logical sense. It is clear that the dialectic tradition, especially in its
ideological (Marxist) version, became guilty of terminological misuse,
with disastrous consequences at the conceptual level: if one gets used
to calling every kind of conflict a contradiction, no attention at all
will be paid to the polysemy of opposites. Instead of differentiating
between contradictories, contraries, and correlatives, the primacy of
contradiction will be established, both in logic and reality. The prin-
ciple of non-contradiction will be criticised as the logical bulwark of
ideology, by means of which the ruling classes attempt to hide social
conflicts. Yes, it must be admitted that there is a great deal of delirium
in this conception.
Bad criticism strengthens the adversary’s position. If separative
thought could be opposed only by “logic of contradiction”, then scis-
sional logic would have no chance. It would not be flexible, warlike
logic, but rather a caricature of reason. So let us hasten to point out
that the relationship between correlatives does not necessarily violate
the principle of non-contradiction: nor is it inspired by this intention.
For example, a particular relationship between master and slave would
only violate Aristotle’s principle if it were and, at the same time, were
not the relationship between a master and slave. No logical transgres-

sion comes to the fore in the overturning of a dominant relationship.
So the relationship between correlatives is not an overcoming, but
rather a version, an interpretation of the principle of non-contradiction;
but if a logical principle is subject to varying interpretations, this
means that it has a stylistic identity. The style of thought interpreting
it is crucial.
Every kind of opposite relationship interprets the principle of non-
contradiction differently. Thus this principle should not be confused
with the relationship between contradictories, which is only one ver-
sion of it. Neither should it be thought (it is perhaps worth repeating
this) that a particular version of the principle, i.e. that of correlatives,
can invalidate, suspend or overcome it. Dialectical extremism is wrong
when it believes it is able to put aside the principle of non-contradic-
tion, but so is separative philosophy, inasmuch as it holds that all ties
between opposites give birth to a contradiction (thus excluding correla-
tives, i.e. the inseparable opposites, from the area of investigation). 36
11. We have taken an important step forward in our research, and

believe it is legitimate to foreground this: when taking a fresh look at
the principle of non-contradiction, we realised that its neutrality – what
makes it absolutely crucial and dominant – does not exclude flexibility.
The separative interpretation is only one of the possible interpretations:
it is the outcome of a fallacy, which consists in superimposing the non-
contradiction principle and contradictory relationship.
The latter statement could be modified, by observing that separa-
tive rationality takes other kinds of relationship into consideration, for
example contrary and subcontrary opposition. This is actually the case,
and there is no doubt that this should be acknowledged. What is essen-
tial is that the difference should be maintained – one could even use the
term ‘heterogeneity’ – between disjunctive and conjunctive relation-
36
Separative thought distrusts ties and interwoven links, but not mixtures and
intermediate cases (this is a point we have already touched upon and to which we
will return).
ships. The principle of non-contradiction presides over all relationships

between opposites37: this does not mean that these relationships belong
to the same logical family.
Family differences became visible by comparing Aristotle’s four
part typology with the medieval logical square: why were privatives/
positives and correlatives left out? While privatives/positives can be
seen as a variant or sub-species of contraries, correlatives contain
something irreducible to the previous types: the enigmatic constraint
linking these terms can only turn out to be incomprehensible to
rationality functioning disjunctively, even when it joins together and
connects (terms remaining separable and potentially autonomous).
The result is the distinction between different logical styles. This need
could be denied, if one were to note that correlatives infringe the prin-
ciple of non-contradiction; in this case, correlatives would be unable to
generate another logic, but only supply separative logic with working
materials – riddles, puzzles, antinomies. However the compatibility
between correlatives and the principle of non-contradiction has been
demonstrated. Correlatives are the most flexible version of this princi-
ple, and, in any case, they reveal its flexibility.
37
Including those concerning future contingents (Lukasiewicz): for example,
“I’ll be at home at midday tomorrow” and “I won’t be at home at midday tomorrow”.
In my view it is trivial, indeed quite wrong, to hold that polyvalent logic deals with
contexts in which the principle of non-contradiction is infringed (Cf Dalla Chiara -
Toraldo di Francia 1999, p. 33). This would only be the case, to return to our example,
if at midday tomorrow I were, and at the same time were not, at home. The indeter-
minacy concerning the future is thus a postponement or suspension, not a violation
of the principle of non-contradiction – a principle in which we have noted surprising
flexibility.
Indeterminism does not appear to infringe the principle of tertium non datur
either, unless we understand this principle as the claim to be able to decide on the
truth of all propositions concerning the future immediately. This claim originates
in a misunderstanding: the possibility of ‘always’ deciding is not the equivalent of
the possibility of deciding ‘now’. These problems deserve further investigation: it is
advisable to be aware of the risk of overvaluing the contribution of polyvalent and
fuzzy logic, trivialising the more recent and difficult perspective: i.e. that of a flex-
ible, scissional logic.
Contradictories are by no means the strongest case of opposition,

as is usually thought: they are the most rigid case. It is quite wrong to
use the rigidity of contradictories as a criterion for judging the other
relationships, and arranging them in an order of ever increasing weak-
ness, dogmatically taking it for granted that all the relationships we
call opposite are, because of the name we give them, homogeneous.
Flexibility is not weakened rigidity, is not the weakest form of rigid-
ity, but something else. Correlatives are not weakened contradictories
but something else. Aristotle’s typology deserves criticism inasmuch
as it does not recognise, but actually tends to hide, the heterogeneity
between rigid and flexible relationships; but the reduction of the logical
square, in both the medieval and Frege’s versions should be even more
severely criticised38.
12. Let us imagine, however, an interlocutor who does not want

to appear dogmatic, and who insists on the homogeneous nature of
opposite relationships; let us try and imagine some of his arguments.
He could mention the words of Catullus, “Odi et amo”, and maintain
that this “et” is a conjunction perfectly accepted and understood by
separative logic, and that there is thus no need to split types of logic
stylistically, positing the hypothesis of a linking logic. How should the
conflictual relationship between love and hate be described? Evidently
not as a case of contradiction – the opposites are mutually compatible
here, albeit in an atmosphere of bitter strife – but as a case of contra-
ries: one of those conflicts between forces to be observed in the natural
and psychic worlds.
This interpretation is possible: but can we consider it satisfactory?
In its apparent neutrality, this description chooses a specific opposite
relationship: that of the contrary, or rather subcontrary. Since an indi-
vidual’s psyche contains a finite quantity of emotional energy, the con-
flict between love and hate, described on the basis of separative logic,
should be transcribed as follows: “I love partially, I hate partially”.
Both statements are true (as can only be the case with subcontraries).
38
G. Frege, Begriffsschrift, 1879.
Among the relationships envisaged in the logical square, that

of subcontraries is the least conflictual, articulated in the weakest
incompatibility39: “some have the y property” – “some do not have
the y property” (or “there is at least one individual who has the y
property” – “there is at least one individual who does not have the y
property”). However, in the case under scrutiny, i.e. the psychic con-
dition described by Catullus, the conflictual condition is especially
intense.
We are somewhat uneasy here: something isn’t working; there is a
false note, as it were, in our tentative analysis. Let us attempt to under-
stand the reasons for this.
13. A person loves and, at the same time, hates another person:
where can we situate this conflict, looking at the square of opposition?
The most plausible answer has already been given: the relationship
between subcontraries. However this relationship asserts the possibil-
ity of an individual denying the universal affirmative proposition, or of
another individual denying the universal negative proposition; it does
not appear to be suitable to point out the case of an individual with
conflictual predicates.
The logical square concerns the relationship between quantifiers,
universal and existential, and that is as far as its function goes. If one
wishes to describe a conflict between predicates in an individual, it is
necessary to take the linguistic-semiotic version of the square into con-
sideration, as suggested by Greimas, for example. Here we do not find
propositions, but rather terms (notions). This construction is entirely
foreign to the dimension of true or false; on the other hand it does add
fresh possibilities: a complex term, generated by the joining of contra-
ries, and a neutral term, generated by the negation of the semes occu-
pying the sites of subcontraries. The carré sémiotique is an abstract
construction, and thus disregards the concrete semantic investment of
39
We are not returning to the traditional point of view. As will be foregrounded
later, in many cases correlatives are more incompatible than contraries and subcon-
traries, although they are, at the same time, bound by a constraint or attraction.
the places it provides for. In the case under discussion, as a matter of

fact, mixed terms are not empty possibilities:
ambivalence
{
love hatred
non hatred non love

{indifference
After shifting the love/hatred relationship to the relevant terrain,
we can hardly avoid newly positing all the previous questions: does the
semiotic square have similar defects to those of the logical square, i.e.
separative style and removal of correlatives? There is no doubt about
this. And the difference between styles can be clarified with reference
to mixed cases.
What Greimas calls a complex term is nothing more than the inter-
mediate case, referred to by Aristotle as possible in the relationship
between contraries40. But is ambivalence a mixture of love and hatred
comparable, for example, to mixing colours, where white and black
produce grey? This seems not to be the case. There are mixtures in
which each opposite does away with its own characteristics, to open
up to a pacified composite. This is not what happens in a feeling of
ambivalence, where opposites do not fade into each other; they actually
engage in furious combat.
At this point, the separative philosopher, or linguist could put for-
ward another proposal: divide, in the area of complex terms, irenic
(more or less harmonious syntheses) from conflictual cases. The love/
hatred ambivalence would then position itself in the latter subset.
How should this proposal be evaluated? It can be seen as partially
plausible: does not everyday thought continually present us with half
truths? Half empty and half full glasses? Nevertheless, the conse-
40
The ambiguity of the examples, in Aristotle, thus turned out to be productive:
it enclosed the possibility of two different versions of the square.
quences of this obstinate inclination to compromise requires some

consideration. Does Odi et amo mean “I hate you to some extent, I love
you to some extent”? Or “I hate you sometimes, on other occasions I
love you”? This description could prove acceptable in cases when this
feeling has cooled down somewhat, only to come to the surface again
intermittently. But the best interpretation of Odi et amo is “I hate
you just because I love you”: the two feelings are opposites, though
they nourish each other, they support each other. Their struggle has
stages when one of the two passions prevails, rather like the case of
Empedocles’ cosmic cycle: absolute love, or absolute hatred. It should
be noted, however, that the triumph of one of the two passions is never
final: in fact, it precedes the resurrection of the opposite passion; thus
giving rise to states of furious mixing. All this happens because the
two passions, although inseparable, are unable to unite. The impossi-
bility of being one is the curse dominating their unbreakable union.
This is indeed the relevant interpretation, if one is referring to really
complex cases, to which one has adequate access. Still in the area of
language: a micro-text such as “Odi et amo” lends itself to simpli-
fied interpretations, and also pointless discussion. It is only when the
context is broadened that one can realise the loss, from the cognitive
viewpoint, caused by recourse to separative connections41.
So the typology of opposites must not be reduced, and the range of
possibilities must always be available. Another mistake to be avoided is
taxonomic rigidity: given two opposite terms, are we able to establish
with definitive certainty what kind of relationship holds them together?
41
Let us re-read Catullus’ text in full: “Odi et amo. Quare id faciam, fortasse
requiris. // Nescio, sed fieri sentio et excrucior” (I hate and I love. Why do I do it, per-
chance you might ask? I don’t know, but I feel it happening to me and I’m burning up).
Other translations are possible, but what counts is the difference between a conflict in
which heterogeneous forces participate (for example, reason and passion) and torture,
nourished by homogeneous opposites. This conflict, like any toxic circularity, can
come to an end, sooner or later: it is not duration that counts but the conflictuality
mode. The difference between contraries and correlatives can be perceived decisively
by observing their dynamism: contraries work on a zero sum basis (for example, the
greater the strength of reason, the lesser that of desire), in correlatives strengthening
is reciprocal (the more I love, and the more I hate and vice-versa).
No, because language is flexible enough to allow twisting and intensi-

fications which only its most rigid structure tends to disallow. Let us
consider the distinction between gradable opposites (like hot-cold) and
ungradable opposites (like male – female). “Ungradable opposites, – as
Lyons observes – when they are employed as predicative expressions
divide the universe of discourse into two complementary subsets (i.e.
the objects of which they are predicable)”42. However, natural languag-
es do not work in a rigid way: one can say that a girl is ‘very feminine’
or ‘not very feminine’, etc. “But in cases like this, we are modifying
the language system, if only temporally. Recognition of the possibility
of grading normally ungradable antonyms, …does not imply that there
is not a sharp distinction to be drawn between gradable and ungradable
antonyms in a language system”43.
Other questions come to the fore: to what extent can a taxonomy
keep its independence, and anteriority, in respect of speakers’ actual
usage? And vice-versa: to what extent does exuberance of usage – i.e.
the pragmatic factor – leave a taxonomy intact?
In that posited by Lyons correlatives are classified as converse:
two place predicative expressions concerning reciprocal social roles
(doctor/patient, master/slave etc.), family relationships (father/son etc.),
temporal and spatial relationships (above/below, in front of/behind,
before/after etc.)44. It would indeed be difficult to justify the argument
developed up to this point on the basis of this classification. Its trivial
nature is partially understandable when taking into account Lyons’
aim: a taxonomy applied to lexemes. There is no opening to the logical
or philosophical dimensions, or even literary analysis, for that matter.
But one has the right to rebel against this humiliating limitation: why
should one acknowledge the legitimacy (or neutrality) of a taxonomy
that entirely ignores a whole area of problems which must be referred
to? These problems are not so far off the mark, seeing that Lyons’ book
has chapters on logic and the philosophy of language.
42
Lyons 1977, p. 271.
43
Ibidem, p. 279.
44
Ibidem, 280.
Confrontation, exchange of ideas between specialists in various

disciplines seems then to be an urgent requirement. As far as cor-
relatives are concerned, their irreducible specificity and abundance
of variants needs to be acknowledged. I will restrict myself to three
research directions:
– the link between correlatives cannot be reduced to a distribution or
sum (whether one considers parts or whole). Solidarity between cor-
relatives is not simply a kind of complementarity, but a very strong
link, reciprocal implication (while the distributive aspect prevails in
complementarity);
– correlatives are often thought statically or as simultaneity:
“Correlatives are commonly held to come into existence together,
and this for the most part is true, for instance, of double and half”,
wrote Aristotle45. This is not the case: the two terms of a correlation
can be consecutive, in accordance with obligatory alternation; think
of toxic objects and the circularity of the two phases of privation and
enjoyment;
– no individual entity can have a contrary, not to speak of a contradic-
tory46. What could be the opposite of Gabriel or Louise? For Aristotle
an individual can hold in him/herself opposite qualities at different
stages, while remaining the same individual, numerically one: a rich
person can become poor, a fat person thin, and vice-versa. Even posit-
ing an exceptional case, when two individuals are entirely character-
ised by diametrically opposite predicates, it is doubtful that one can
speak of a contrary relationship.
Two individuals can be friends or enemies etc., but not contraries. A
correlation could be set up between them, however; and this is not an
exceptional, almost unthinkable case, but rather easily and frequently
encountered reality. This is another reason for not removing correla-
Aristotle, Categories 7, 7 b 15;

45
“Substances never have contraries. How could first substances have them - this
46
man, for example, that animal?” Aristotle, ibidem, 5, 3b 24-25.

tives from Aristotle’s typology: they are the only case of opposition
concerning individuals, and not predicates or quantifiers.
14. Correlatives are the obstacle or rather adversary in the way of

separative thought, revealing its rigidity; so every effort is made to
dissolve the specificity of this relationship or resolve it within a domes-
ticated taxonomy. By claiming specificity for correlatives, however, a
more accurate typology is not merely aimed at – this objective would
be pretty poor: the specificity of correlatives is absolutely crucial
because it is the possible source of non-separative logic.
Let us return to the principle of non-contradiction. Linking it with
the typology of opposites, with its variety of types, we have shown its
stylistic identity: the principle of non-contradiction is always implicitly
modalised by a style of thought. It can act in the sphere of rigid or flex-
ible thought: this set of possibilities implies that flexibility resides at
the very heart of this principle, the rigid version of which has almost
always been enhanced.
Can the difference between versions influence the formulation
modes? According to Aristotle’s formulation, the principle of non-
contradiction states that it is impossible for the same attribute at once
to belong and not to belong to the same thing and in the same relation47.
There is another formulation, generally accepted as equivalent, i.e. the
principle of identity, “A = A”. But are these two formulations really
substantially equivalent? Heidegger’s answer is negative. In his view
something happens in the passage from the traditional formulation to
the one making use of the ‘equals’ sign: the polysemy of the verb to be
is eliminated in favour of a univocal indication.
Polysemy is not a defect, a gap, a source of disorder and confusion,
but rather a set of possibilities. Admittedly, the polysemy of a concept
can be reduced to a list, in the sphere of which each meaning lines up
with the others in mutual indifference: actually, this is the way separa-
47
Aristotele, Metaphysics, IV 1005 b 20. The reference to point of view is often
neglected as obvious: if I say Alex is taller than Paul and, at the same time, shorter
than Jim, I am not falling into a contradiction.
tive philosophy operates, and not only in modern times. In the case of
the verb to be, the most fascinating one, but also the vaguest in philo-
sophical language, in order to dissolve its aura and abolish its mysteri-
ous reserves of meaning, it would be enough to distinguish between
the copula function, that of identity and that of existence. “Thus the
word ‘is’ figures as the copula, as a sign for identity, and as an expres-
sion for existence”48. Naturally conjunctive philosophy does not reject
these distinctions, but believes that it is indispensable to analyse a set-
up of distinctions (list, taxonymy, typology) from the stylistic point of
view. Thus any polysemy potentially splits into modes of articulation:
there are typologies with a well ordered, homogeneous manifoldness,
with no internal disputes, while others, and we have seen this in the
case of opposite relationships, are the product of a separative fallacy
which attempts to cancel the heterogeneity of types. Therefore, the
conflict permeating them must be rediscovered.
If Heidegger posits retranslating “A = A” by “A is A”, this is so as
to return to a hastily investigated polysemy, which is wrongly believed
to have been completely mastered, at least without possible surprises.
Returning to the verb ‘to be’ allows access to the problem of ontologi-
cal difference, but also allows one to think of the connection marked
by “is” not as identity/equality or existence, or belonging, but rather
Zusammengehörigkeit (belonging together). Heidegger frequently, and
always at crucial stages in his arguments, addresses this notion, which
refers to reciprocity, mutual inclusion, and therefore correlation. We
cannot deal with the working of Zusammengehörigkeit in Heidegger
here, so we shall restrict our attention to some passages concerning the
problem of identity.
The departure point is scission of synonymy. Heidegger rejects
equivalence between equality or identity (das Gleiche) and sameness
(das Selbe): “In the merely identical the difference disappears. In the
48
Wittgenstein, in Tractatus, 3.323. The meaning of the copula can be further
subdivided into two types: inclusion of classes (“Every Frenchman is jovial”) and
an element’s belonging to a class (“Abelard is French”). The following sentences are
examples of identity and existence “Aldous is the King of France” and “There is at
least one Frenchman”.
same the difference appears,…”49. This means that das Selbe does
not exclude but includes difference: the concept of “sameness” in
Heidegger is interpreted conjunctively. Let us attempt to clarify this
position further by translating it into our own language.
The identical, in the sense of ‘equal’, excludes the non-identical.
The two concepts are incompatible, whether thought as contradictories
or contraries: rigidly incompatible, as claimed by separative thought.
Heidegger, on the other hand, argues that a relationship in which oppo-
sition between identical and non-identical is inclusive can be thought.
What could we call this relationship? Do we already have a name, and a
conceptual elaboration, able to refer to it and describe it correctly? This
appears to be the case: if the identical and non-identical belong together,
it is because they are correlatives. Heidegger’s Zusammengehörigleit is
an interpretation of Aristotle’s correlatives.
There is no doubt that there is something new, however. Aristotle’s
typology has a ‘constative’ character: it says that cases of correlation
exist. However, we have already seen that any typology can be inter-
preted dynamically, thanks to the inexhaustible flexibility of language.
Non-gradable antonyms, such as male/female, can be treated as grada-
ble and appear in sentences that make perfect sense. Logic can also act
flexibly, turning opposition between contradictories or contraries into
opposition between correlatives. The problem is: is this always possible
in principle?
15. The scission of the synonymy between das Gleiche and das
Selbe, in Heidegger’s perspective, makes the division of the notion of
‘identity’ possible. Identity does not necessarily mean ‘coincidence’: it
can also be understood as ‘non coincidence’.
Merely defining identity as coincidence is thus a fallacy. As far as
we are concerned, we will no longer say that “A = A” is the principle
of identity, unless we add that it is the principle of separative identity,
i.e. the separative version of the identical.
49
Heidegger 1957, p. 45.
Let us return to the question so far unanswered: is it always possible

to treat relationships between contradictories or contraries as relation-
ships between correlatives? That is to say: is the principle of non-coin-
cidence – this is what we can call the scissional-conjunctive version of
the principle of identity – referable and applicable to any entity? We
must look again at the meaning of correlation, in identity, between the
identical and non-identical. This conception of identity could be seen
as bizarre and incomprehensible: how can one deny that everything is
identical to itself and nothing more? Does not identity, strictly speak-
ing, i.e. the relationship between an entity and itself, necessarily have a
separative meaning? Everything is separate from everything else, and
is inseparable from itself.
So, the scissional identity thesis (hinging on the correlation between
identical and non-identical) strikes a double target: it denies the insepa-
rability of a thing from itself when denying its separability from other
things. Identity is a relationship, but not necessarily the relationship
separating an individual from others; it is rather – for a specific type
of entity – the relationship joining him/her to some of them. Identity is
thus understood as ‘identification’.
16. We use this term in the psychoanalytical sense. In Freud’s view,

the identity of every human being coincides with the series of his/her
identifications: identification is seen as a wide ranging, often conflict-
ual, mostly unconscious process significantly modifying the previous
personality. There is no Self preceding some process of identification.
The relationship between the Self and the Other is therefore essential
and constitutive. Unawareness of these processes and believing oneself
to be a fully autonomous subject is part of the mendacious character
of the Ego. Note that the identification process is asymmetrical: A
becomes B, partially interiorising him/her, but B does not become A50.
50
This relationship could be symmetrical: 1) if A identifies him/herself with B
and B identifies him/herself with A; 2) if A, B, C, etc., identify themselves with the
same individual (for example a leader) and, as a consequence, all identify themselves
with each other. This is the case described by Freud in Psychology of the Masses and
Ego Analysis, 1921.
The concept of identification upsets the classic theory of predica-

tion, from Aristotle to Frege, Quine, etc. On the basis of the distinction
between singular terms (like Socrates) and general ones (like Athenian,
mortal, etc.), it is usually stated that a singular term is susceptible to
predication (I can say that “Socrates is an Athenian”) but cannot occur
as a predicate (I cannot say “Plato is Socrates”). For conjunctive logic,
on the other hand, singular terms can also stand in the predicative
position. On the basis of the theory of identity as identification, it is
quite legitimate to say that Plato is (was) Socrates: by this we mean,
and it cannot easily be challenged, that Plato identified himself with
his teacher, took up his teaching and developed a philosophy inspired
by his thought. Admittedly, we must distinguish the types and modes
of identification, which is a scissional-conjunctive process: identifica-
tion can be partial or almost total, can refer to the Ego or the Ideal of
the Ego, can be distinctive or confusive (a confusive example is Don
Quixote’s identification with Amadis of Gaul).
It should be pointed out, however, that a process of identification
is characterised by a complexity that can only be partially described
by means of predicative connections. When identifying him/herself
with B, A assimilates some of his/her traits, and they can be indicated
through properties or behavioural modes: it can be said, for example,
that A is (has become) “a knight-errant defending Christianity, fight-
ing against injustice etc.” but what would this description be worth as
far as Don Quixote is concerned? Even if we were to list all the traits
Cervantes’ character had absorbed from his model, the description
would remain schematic and insufficient51. It is interesting to give
these difficulties some thought, because it is from here that certain
aspects of the problem under investigation can be clarified.
51
The initial pages of the novel describe these traits: “He believed that it was
necessary, both for his own honor and for service of the state, that he should become a
knight-errant, roaming through the world with his horse and armor in quest of adven-
tures and practising all that had been performed by the knights-errant of whom he had
read. He would follow their life, redressing all manner of wrongs …”, etc. (Cervantes,
Don Quixote, Part I, Chapter I, 1616, p. 59). This is clearly not enough to describe the
personality of the leading character.
17. At this stage we must ask ourselves what the limits to be

assigned to the new theory of predication are. We have established the
fact that a singular term can appear – without any incoherence, and
making perfect sense – in the predicate position (“Plato is Socrates”,
“Don Quixote is Amadis of Gaul”); but this does not mean that any
singular term has the right to take on a predicative role and propose
itself as the identification term.
For example, could one refer to a poplar which is – which has
identified itself with – an oak tree? The sentence seems bizarre, not
only because semantic rules or classification spheres are broken, but
because the desire to be evidently only concerns entities equipped with
consciousness. But even if we were to allow for an embryo of con-
sciousness in trees, the metamorphosis would turn out to be implausi-
ble: metamorphosis – what else is identification? – is a process that can
only concern flexible entities, entities whose identity is not reflected
in the “A = A” formula, but in the correlation between identical and
non-identical. It is only for flexible individualities (here there is no
need to make a list: we are discussing logic) that the principle of non-
coincidence is valid.
Two points:
(a) it is important not to overlook the difference between a process of pro-
perty substitution (Lewis was rich, nice, smart etc.; he has become poor,
nasty, scruffy etc.) and one of metamorphosis linked with identification.
When predicates, by means of which an individual has been described,
are replaced by a large number of opposite predicates, the individual can
be said to have drastically changed, to have undergone a ‘metamorphosis’.
However, Lewis could have turned nasty because he is now financially
worse off, which prevents him from taking care of his outward appear-
ance, etc., without this upset having anything to do with identification
processes: actually, identification is not a ‘property’ metamorphosis,
which can be described by means of an alternation of predicates. There is
a perspective component in this process: identifying oneself means tak-
ing up the other’s gaze, his/her vision, values and taste;
(b) we have already mentioned the fact that the principle of non-
coincidence only works in the sphere of the flexible; an entity with a
rigid identity is situated, on the other hand, in the sphere of separative

identity, A = A. Metamorphoses can also be imagined here: just think
of Escher’s engravings, for example, the strange triangle which looks
as if it has absorbed the traits of some other geometric figure. This is
the point, however: we are considering a combination with heuristic,
cognitive value, which can be judged true or false, as in the case of the
sentence “Plato is Socrates”, and not the result of unruly morphologi-
cal imagination? Certainly, this imagination can produce extremely
fascinating effects, belonging to the confusive; we cannot address the
problem here of the possibile truths of the confusive regime.
18. There is a further question: is not the border between rigid and
flexible in turn flexible? This is not the equivalent of the absence of a
borderline, a week or vague border, though it forces one to ask oneself
about the possibility of metamorphosis. We will now quickly deal, in
the small amount of space still available, with metamorphoses in lan-
guage, i.e. in the modalities of language named figural.
The most fascinating rhetorical figures (metaphor, oxymoron, etc.)
are actually conjunctive. In figural language we encounter metamor-
phoses of identity recalling processes of identification; an affinity
confirmed by the fact that Lacan indicated the identification relation-
ship of a son with his father, and subsequent access to the Symbolic,
with the expression paternal metaphor: the father acts as a modelising
force, similarly to metaphorical focus (if we turn to Black’s theory). He
selects, foregrounds and transforms.
Differences must not be neglected, of course: identification, from
the point of view of psychoanalysis, is a process whose protagonists
are generally perspective entities. This does not prevent us from think-
ing of cases of identification with an abstract entity (an institution: the
University, the State etc.) or even concrete objects52. Rhetorical-figural
processes are based, on the other hand, on relationships established by
subjects imposing their perspective on entities that may not have one
52
“O that I were a glove upon that hand, // That I might touch that cheek” (W.
Shakespeare, Romeo and Juliet, 1595, II, 2, vv. 24-25).
(or in the case of only one of the entities having one, as with “Joe is a
flash of lightning”).
The analogy between psychic and linguistic metamorphoses appears
to break off completely if one takes on the point of view of predication.
We have seen that, in the case of identification, a singular term can be
the predicate of another singular term; rhetorical figures like metaphor
appear to be closest to classic predication, even when they place an
object, rather than a quality in the predicate position. When saying “Joe
is a flash of lightning” we are not referring to any particular flash of
lightning, i.e. an entity identified in space and time, as when one says
that Plato is Socrates, but rather to flashes of lightning in general, the
flash of lightning class, defined by specific properties.
Nevertheless, there is a lingering feeling that a metaphor describes
a non-property metamorphosis, one that is not entirely translatable
into a series of predicates. Predicative paraphrase implies cognitive
loss: what is lost is the perspective component, which is grounded in
the individual positing or inventing the metaphor, and which can also
be perceived in the metaphorical utterance; the foregrounding carried
out by the metaphor is not perceived, and this is a function frequently
neglected by scholars restricting their interest to the similarity connec-
tions: metaphor partially acts as hyperbole, intensifying and distorting.
If these aspects are taken into account, metaphor tends to move away
from the classic predicative scheme.
Moreover, have we not perhaps learnt from modern philosophy of
language to mistrust the exterior appearance of a sentence? We know
that there can be a great distance between grammatical and logical
form. So why should we be inhibited by the fact that the metaphoriser
appears as a general term? What counts is its logical form, and, from
the logical point of view, the metaphorising expression (the focus of the
metaphor) appears to act as a singular term, or position itself half way
between classic and singularity predication.
This hypothesis is further confirmed if we look at oxymorons.
Irrelevant cases should be avoided, i.e. those where the conflict is not
true opposition. This happens, for example, when the conflict does
not concern the same but different parts of the psyche (leaving aside
adherence to Freud’s theory of the divided subject, one can continue to
use traditional psychology of faculties). Thus Zerlina’s “Vorrei e non

vorrei” in Mozart’s Don Giovanni (I, 9) is not really a case of an oxy-
moron, since it originates in two different areas of the psyche, desire
(“felice inver sarei”) and reason, the rational fear of being deceived
(“ma può burlarmi ancor”). Here the conjunction “e” (and) is not oxy-
moric or paradoxical, but rather antithetic and separative. An excellent,
paradigmatic case can be found, on the other hand, in the phrase “the
light of darkness”53.
The criterion for discovering authentic oxymorons could be the
semiotic square (whose limits we will provisionally accept). While the
conflict between desire and fear is not paradoxical opposition, since the
notion of ‘desire’ is the contrary of ‘repulsion’ and the contradictory of
‘non-desire’, light and darkness can make up a paradoxical link.
In the phrase under examination, however, opposition has been
transformed into specification: what light are we talking about? Light
attributed to darkness. There is no antithesis, but rather a predicative
connection: light is predicated by darkness.
Incompatibility between opposites has been overcome; disjunction
has become conjunction. Once more we can take note of the flexibility
of language. But what kind of link has been set up? Incompatibility
does not appear to have been removed by means of an intermediate
term (or complex one, as Greimas calls it); the mix between light and
darkness is arguably intermediate, as perceived at dawn and sunset.
The semantic intention giving rise to this phrase can be recognised
if we translate it into a perspective: try to imagine glaring darkness,
which can dazzle with the same strength as light. Dazzling, blinding
darkness.
Have we obtained a classic predicative form? Or a conjunctive
transformation? We have not mixed opposites – the only possibility
accepted by separative logic, i.e. mixed cases – , we have tied them
together. But this is not an aporetic node: it does not produce paralysis,
as is the case when I say that something true is false or something false
53
From Gracián, Oraculo manual, 1647 (section 13). In a strategic context intelli-
gence “discovers the darkness concealed by the light and deciphers every move,…”.
is true54. Here opposites strengthen one another – is this not what hap-
pens in Heidegger’s ‘belonging together’ (Zusammengehörigkeit) ? We
cannot go further into the resemblances between different nodes, and
are unable, here, to explain why some oppositions produce paralysis
while others do not. Perhaps, in all non aporetic correlations, there is
an asymmetrical component. “Festina lente” (Make haste slowly)55:
another excellent example of an oxymoron. This is not an invitation
to the intermediate behaviour (be neither too quick nor too slow).
The golden mean (mesotes) 56 is not being pointed out here, but rather
a ‘golden extreme’: hurry up – speed is a strategic virtue –, without
giving up the advantages of taking your time. It is not a suggestion in
favour of the intermediate condition, but one of inclusion: to take it over
it is not enough to calculate; shrewd, strategic intelligence is required.
Augustus addressed this maxim to his commanding officers.
19. Here follows a list of theses, considering the route followed so

far and its possible developments:
– scissional (conjunctive) logic is paradoxical, but not aporetic logic;
– if the scissional component is missing or fails, conjunctive logic becomes
confusive. Opposites are tied together, as it were, by over tight knots,
which give rise to a block or paralysis. Movement is no longer possible;
54
A sensible interpretation can also be given for these utterances (or at least one
of them): this is the case when “Sagacity…tries to deceive by truth itself” (Gracián
1647, ibidem). Recall the story told by Freud: “Two Jews met in a train in a station
in Galicia. ‘Where are you going?’ asked one. ‘To Cracow’, was the answer. ‘What a
liar you are’ broke out the other. ‘If you say you’re going to Crakow, you want me to
believe you’re going to Lemberg. But I know that in fact you’re going to Cracow. So
why are you lying to me?” (Freud 1905, p. 115).
We could sum up this as follows: “What I am saying is true. Therefore it is false”.
Is this a new paradigmatic case? An alternative to the Liar paradox? The link between
true and false illustrated in this short narrative confirms the need for an investigation
that is not guided by separative rationality.
55
This maxim is attributed by Suetonius to the Emperor Augustus (Life of
Augustus, 25, 4).
56
“Virtue, therefore, is a mean state in the sense that it is able to hit the mean”
(Aristotle, The Nicomachean Ethics, 1106 b 28).
– ancient (Aristotle) and modern (Frege, etc.) logic turned to a single

logical style, i.e. the separative one: a style that should reduce the possi-
bilities of paradoxes to a minimum, or at least ensure their solution57;
– the separative style is perhaps able to unlock all the enchained para-
doxes or aporias; but cannot think the logical ties called correlatives
by Aristotle, which are the source (or principle) of flexible logic. Ties
between correlatives are not errors or confusion, except for the case of
the subset of aporias;
– the postulate of incompatibility between opposites must be elimi-
nated (just as the postulate of parallels was denied in geometry): it
prevents the onset of flexible, scissional logic;
– the postulate of incompatibility has its best known expression in the
square of opposites (logical version). The linguistic version (semiotic
square) allows for mixed cases, albeit thought as derivations. Both ver-
sions of the square are subject to the separative;
– a typology of opposites must not ignore or minimise the heterogenei-
ty between disjunctive and conjunctive relationships;
– neither should one ignore or minimise the difference between iden-
tity as coincidence and identity as non-coincidence (scissional, correla-
tive identity);
– flexible logic is quite different (unless proved otherwise) from poly-
valent logic, and also fuzzy logic. It is not enough to admit (or gener-
ate) unprovable propositions to create flexibility;
57
In this essay we have encountered a number of attempts at escaping from the
rigidity and aridity of the logic launched by Frege: the escape route was mostly looked for
in open-endedness and vagueness (many-valued and fuzzy logic). We did not consider
another important attempt – this time by Graham Priest and paraconsistent logics.
We can, however, at this stage indicate the point beyond which there is no access
for all these searches and the reason why they seem so disappointing: none of them is
able to posit the problem of pluralism among styles of thought, and acknowledge the
monostylistic fallacy they have in common.
I do not believe that I am being unjust when stating that the logic of flexibility is
of a radically different kind. It is another logic: a mental space which no deviant log-
ics – many-valued, fuzzy or paraconsistent – can in any way even imagine.
– Freud’s principle of anaclisis must be conceptualised as a logical

principle;
– the list of logical connectives must be newly expressed and expand-
ed: in logic that is no longer ‘zero sylistic’, a priority is the introduction
of the connective as;
– rhetorical figures, or at least some of them, such as metaphor and
oxymoron, correspond to conjunctive mechanisms; beyond grammati-
cal expressions, they should be studied in their logical form.
– what we call language is a mixture and conflict of styles.
APPENDIX
THE CURVED LOGIC OF CORRELATIVES:

A REFORMULATION OF THE SQUARE OF OPPOSITION
Is it possible to reformulate and expand the logical square, by

introducing conjunctive relationships? Or must we consider it a tool
only capable of articulating disjunctive relationships (between contra-
dictories, contraries and subcontraries)? Even if we reached the latter
conclusion, there would be no reason for disquiet on my part. In my
view, what counts is having stated the legitimacy of a perspective to be
understood and judged above all conceptually (technical developments
follow on from this).
I have tried to remove the prejudice, according to which there is
supposed to be a preferential link (an affinity) between the principle of
non-contradiction and disjunctive relationships. Actually, the principle
of non-contradiction exists in its different versions, and correlatives
are the most flexible of them. Consequently, inasmuch as it excludes
conjunctive relationships, denying them logical legitimacy, the square
of opposition is a major fallacy, and one must rid oneself of this fallacy
once and for all.
How can this be done? By acknowledging the limits of the logi-
cal square (its monostylism) and trying out a richer, more complete
elaboration, even graphically. But how can correlatives be represented
graphically? This is my suggestion:
The two arrows pointing at each other represent conflict (I am not

interested in eirenic correlatives); the curved line connecting them rep-
resents interdependence, reciprocal, necessary presupposition
Where can the relationship between correlatives be inserted?
There are two possibilities: ‘horizontal’ insertion, above the relation-
ship between contraries, and a ‘diagonal’ one, next to the relationship
between contradictories. Here it is enough to illustrate one of the two

options, the one accentuating paradoxical characteristics:
But the curve was principally chosen so as to indicate the elasticity,

the flexibility of correlatives in comparison with the rigid linearity of
the contradictories and contraries. One further, albeit obvious, point:
when a logic is curved it is so in a way unlike that of rails or a road, or
a jar handle (because these forms belong to rigidity).
Bibliography.
Arduini 2007 Metaphors Concepts Cognition in Metaphors, ed. Stefano
Arduini, Roma, Edizioni di Storia e Letteratura.
Bottiroli 1997 G. Bottiroli, Teoria dello stile, Firenze, La Nuova
Italia.
Bottiroli 2007 Metaphors and Modal Mixtures, in Metaphors, ed. Ste-
fano Arduini, Roma, Edizioni di Storia e Letteratura.
Clark 2002 M. Clark, Paradoxes from A-Z, London, Routledge,
2002.
Dalla Chiara – Toraldo M.L. Dalla Chiara - G. Toraldo di Francia, Introduzione
di Francia 1999 alla filosofia della scienza, Roma-Bari, Laterza.
Engel 1998 P. Engel, La vérité. Réflexions sur quelques truismes,
Hatier, Paris.
Frege 1915 G. Frege, Meine grundlegenden logischen Einsichten,
1915, in Nachgelassene Schriften. English translation by
P. Long and R. White, Oxford, Blackwell.
Freud 1905 S. Freud, Der Witz und seine Beziehung zum
Unbewussten, 1905; English translation, Jokes and their
Relation to the Unconscious, 1960.
Freud 1915-1917 S. Freud, Vorlesungen zur Einführung in die Psychoanalyse,
1915-1917; Introductory Lectures on Psychoanalysis (1915-
1917), English translation by James Strachey, The Hogarth
Press, London 1963.
Haack 1978 S. Haack, Philosophy of Logics, Cambridge, Cambridge

University Press, 1978.
Heidegger 1957 M. Heidegger, Identität und Differenz, 1957; Identity
and Difference, translated by J. Stambaugh, New York,
Harper & Row, 1969.
Klein 1991 E. Klein, Conversations avec le Spinx, Paris 1991, English
translation by David La Vay, London, Souvenir Press.
Kosko 1993 B. Kosko, Fuzzy Thinking: The New Science of Fuzzy
Logic, London, Harper Collins.
Kripke 1975 S. Kripke, Outline of a theory of truth, «Journal of
Philosophy», 72 (1975), pp. 690-716.
Lyons 1977 J. Lyons, Semantics:1, Cambridge, Cambridge University
Press.
Nietzsche 1883-1885 F. Nietzsche, Also sprach Zarathustra, 1883-1885;
English translation by Thomas Common, Ware,
Wordsworth.
Odifreddi 2001 P. Odifreddi, C’era una volta un paradosso, Torino,
Einaudi.
Penco 2002 C. Penco, Introduzione, in La svolta contestuale, Milan/
New York, McGraw-Hill.
Penco 2005 C. Penco, Introduzione alla filosofia del linguaggio,
Roma-Bari, Laterza.
Popper 1960 K. Popper, On the sources of knowledge and ignorance,
1960, now in Conjectures and Refutations, Routledge &
Kegan Paul, London 1963.
Tarski 1952 A. Tarski, The semantic conception of truth and the
foundation of semantics, in Leonard Linsky, Semantics
and the Philosophy of Language, University of Illinois
Press.
Varzi 2001 A. Varzi, Parole, oggetti, eventi e altri argomenti di
metafisica, Roma, Carocci.
Wittgenstein 1953 L. Wittgenstein, Philosophische Untersuchungen, 1953;
English translation Philosophical Investigations, by M.
Anscombe, Blackwell, Oxford.
Wittgenstein 1977 L. Wittgenstein, Vermischte Bemerkungen, ed. G.H. von
Wright, 1977; English translation Culture and Value by
Peter Winch, Blackwell, Oxford.
PETER CARRAVETTA
NO LONGER A PARADOX: THE SOPHISTS

AS PHILOSOPHERS OF LANGUAGE AND EXISTENCE*
In the fifth century before the Common Era there appeared in the
great city of Athens a group of progressive, “liberal” intellectuals called
the Sophists. They were by and large foreigners, not exactly barbaros, but
more like metics, resident aliens who, insofar as they might have gained
citizenship, could claim some civil rights, though they were not allowed
to participate in city politics. The sophists were professional teachers,
specializing in lecturing on a variety of subjects and above all on lan-
guage itself. For the first time in history, these intellectuals were paid for
their services. They met a specific sociocultural need. In the sixth century
Solon had effected some liberalizing reforms in the polis, and at the end of
that century Cleisthenes, an archon or magistrate in Athens, introduced
voting privileges, which some have argued signals the true beginning of
democracy in the West. Not much later, Ephialtes and Pericles introduced
further changes which basically transferred power from the Aeropagus to
the Assembly. This meant so many more citizens could aspire to partici-
pate in running the commonwealth, but also, and just as crucial, advance
in social class and status. These new opportunities created a demand
for instructors in the arts of debate who could moreover furnish a well-
rounded education. The sophists taught also areté, usually translated as
virtue, but as translation is ever and always also an interpretation, things
*
This paper is a slightly modified chapter from my forthcoming book, The
Elusive Hermes. Method, Discourse, and the Critique of Interpretation, Aurora (CO),
Davies Group Publishers.
62 PETER CARRAVETTA
are much more complicated. Insofar as it was meant to equip this new
burgeoning class with informed agendas and effective tools to partici-
pate in domestic public and social life, it had the sense of what today in
America we call “values”. No need to remind the reader how discussions
about values are today tossed about by politicians, educators, parents and
the media, that it is practically impossible to come up with a neat defini-
tion of what values really are, let alone comparing them to those of other
countries. Nevertheless by a cautious retrojection of the criss-crossing of
cherished, contested, vilifying or misunderstood accounts of what a given
value may have entailed or required in the fifth century, we can in part
surmise what the real world of the Sophists was like. At least, glimpse
their interpretive horizon. Already in Athens we have a broader spectrum
in social classes, competing interests, and the thrust of a protobourgeoisie
of sorts. New needs arise, and possessing noticeable speaking ability was
a key factor in this changing public life.
But it was not only speech the sophists taught, for there is no dis-
course without some kind of intention and content. The sophists were
aware of this, they consciously taught ideas as well, some of which
were very paradoxical by our post-Platonic, post-Aristotlean ethical and
mental habits. Some, indeed, were downright radical – we actually have
in the surviving writings claims which in terms of our recent historical
memory would sound like “all men are created equal”, and even “God is
dead”! Above all, they taught that through language one can be empow-
ered, may even gain access to the means of production and political
authority. And that was bound to meet with some reticence, at times
resistance, often explicit rejection. When we citizens of the institution-
alized golden mean read about this state of affairs, from the plateaus of
the XIX and XX century, we reflexively when not indignantly point out
that these here sophists dared to teach for a fee, when ideally friend-
ship and gratitude should have sufficed. But whose “ideal” was it, this
of free education? Who could afford an education? And is it not high
time we revise the lofty-sounding and retroactively imposed mythology
whereby, at least in the West, knowledge should be pursued for its own
end, and that learning, wisdom, is not something to be traded? Parents
may teach their children for free, but for the past half a millennium any
imparting of knowledge has been professionalized, and it is absurd and
PARADOXES: THE SOPHISTS AS PHILOSOPHERS OF LANGUAGE AND EXISTENCE 63
hypocritical to cite the fact that the Sophists “charged a fee” in quota-
tion marks as if to suggest, my goodness, how could they do that? They
must have been in bad faith! Well, educators and professionals on a sal-
ary should not be irked or ironical about this.
The real problem, perhaps, was another, for “the sophists sell
wisdom to all comers without discrimination” (Kerferd 1999, p. 25).
That egalitarianism, and the fact that they abjured, rejected or ignored
notions of a Supreme Being and Epistemological Unity, set the stage
for the negative and often dismissive interpretations of their message
and thereby any possible contribution to philosophy1. In the twentieth-
century, in Euroamerica, the Sophists would be right at home, if only
one would connect them to such names as Albert Einstein, Werner
Heisenberg, Paul Feyerabend, Jacques Derrida, Marshall McLuhan,
chaos theory, the American Bar Association, the no longer serious
issue of truth in advertising, the ubiquity of political infomercials, and
the reckless economic law of the market and its “ethical” to wreck
people’s lives justification, among other things.
The challenge that Protagoras submits to the Eleatics is that neither
God nor Being, but “man is the measure of all things”. This famous
1
This trend of course begins with Plato, who features sophists in several of his
dialogues, and gets progressively worse through the ages. A partial reappraisal began
with Hegel, who “re-introduced” them into the history of philosophy as representa-
tive of the “subjectivists”. The historical reconstructions effected by Grote, Zellner,
Nestle and Guthrie, though meant to validate the sophists’ contribution to philosophy,
pedagogy and Greek thought in general, are dotted with caveats similar to those
we find with the scholars of Greek myth. Cf. Kerferd 1999, pp. 4-14. It is largely
owing to the work of Untersteiner, Kerferd and Schiappa that the sophists have been
reintroduced into contemporary discussion outside of the closed circuit of classical
philology and into philosophy and language studies, making over nearly half century
a solid case for their rehabilitation. And it is only in very recent times that some even
more far ranging revisionist work has been carried out; cf. in particular listed works
by Pullman, Poulakos, Enos, Scenters-Zapico, and Bett. Scott Consigny identifies
two camps in this rehabilitation movement, the “foundationalists”, which includes
Eric Havelock, Jacqueline de Romilly, Kerferd, Edward Schiappa, and Thomas Cole;
and the “anti-foundationalists,” or neosophists, which would include, among others,
Sharon Crowley, Victor Vitanza, Katheleen Welch, and Susan Jarratt (253). A detailed
study of these currents will appear in a separate study.
64 PETER CARRAVETTA
assertion is at the root of much wrangling among philosophers, histo-

rians and philologists, most of it owed to the fact that each succeeding
generation interprets the previous one in terms of what suits its ideo-
logical bent and intellectual agenda. In principle, there is nothing wrong
with this, as it is inevitable and by and large explainable: hermeneutics
requires also that we consider the history of interpretations correct or
inappropriate as they might have been. What is relevant is how this is
carried out and with what consequences – often planned, but even more
often unintended, – on the later social and cultural lifeworlds which
make recourse to a particular authority to explain itself to its constitu-
ency. The key passage reads:
For [Protagoras] says somewhere that of all things the measure is man, of
things that they are that they are, and of things that are not that they are not2.
Suddenly, it is humans who find themselves with the onus of
accounting not only for the godhead, not only with aleatory Being, but
for themselves, for their existence. Pneuma begins its long travailed
course toward psyche, then toward spirit, someday, in the Modern
epoch, it will identify itself as ratio. But, and no pun intended, ratio
has had to deal all along the route with oratio. Long before the aware-
ness of causal relationships – I think, therefore I am, – the sense of
a split between supreme power and personal capacity is introduced,
but unlike what happens with the Eleatics, the focus is no longer the
Unsayable, but rather what can be said, and what can, indeed what
must, be articulated is that the reference point is now human existence
itself. Before the preposition which can guide meaning, there is a new
meaning that attracts to itself everything in the cosmos and everything
on the earth: I exist, people exist, things exist: now what?
How do we link these disjointed generators of meaning? And what
of the disclosing of a social world with a plethora of new possibilities,
2
Plato, Theaetetus, 152A, which Diels-Kranz consider an authentic fragment
directly intercalated in the dialogue. I am citing from Sprague 2001, p. 19. Cornford
translated it thus: “[Protagoras] says, you will remember, that ‘man is the measure of
all things – alike of the being of things that are and of the not-being of things that are
not’” (in Hamilton & Cairns 1978, p. 856).
some of which later generations preferred calling, in part to reassure

themselves by ducking under formal systems, contradictions? Things
are, and things are not: I can only know what I see, what appears to
me. What does that mean? Western philosophy usually picks this issue
up beginning with Plato, who struggles mightly to tie the anarchic ten-
drils together, both in the Theaetetus and in the Protagoras. The above
citation sets the stage for what amount to the first misrepresentation of
the sophist:
Now doesn’t [Protagoras] say something of this sort, that as each thing
appears to me, so it is for me, and as it appears to you, so in turn it is for you,
you being a man, and I too?...Isn’t it true that at times, when the same wind
is blowing, one of us will be cold and the other will not, or the one slightly
and the other extremely so? –Indeed it is.– Now in that case shall we say that
the wind is cold in itself or not cold, or shall we agree with Protagoras that
it is cold to the man who feels cold but not so to the other? –We shall agree
with him, it seems. –It also “appears’ so to each one, doesn’t it?– Yes.– And
‘it appears’ surely is the same as ‘he perceives’? –It is. – Appearance, then,
and perception are equivalent when one is speaking of warmth or anything
of that sort. Then things are, I venture, for each person just as he perceives
them (Theaetetus 152a-d)3.
One can sense how Theaetetus is being “set up” as an accomplice
in the ensuing sarcastic character assassination4, in order for Socrates,
well, for Plato, to pursue his discrediting of the sophists, and continue
on his quest of re-instating a supreme order, the Theoros, clearly an
authoritarian one, which for half a century had nearly disappeared.
3
Cornford’s translation is clearly vitiated by his desire to make Socrates’ attack
more readable and acceptable to XX century minds sold on the proverbial Socratic
pursuit of knowledge. The last response above is as follows: “Perception, then, is
always of something that is, and, as being knowledge, it is infallible” (in Hamilton &
Cairns 1978, p. 857). Here translation equals interpretation as rationalist ideology.
4
This will occur in part at 161c: “On the whole I’m quite delighted with his state-
ment that what appears to each man also is. But I am surprised at the way he started
his account, that he didn’t say at the beginning of his Truth that of all things the
measure his the pig or the baboon or some even more outlandish choice from among
creatures endowed with sensation… he was in fact no more intellegent than a tadpole,
to say nothing of other men” (Sprague 2001, p. 19).
66 PETER CARRAVETTA
What had the sophists wrought? A few lines from Antiphon, who
has been cited as affirming “God has no need of us”, (DK B80) should
suffice:
We [respect] and revere those who are of good parentage, but those who are
not of good family we neither [respect] nor revere. In this behavior we have
become like barbarians to one another, when in fact by nature we all have
the same nature in the particulars, barbarians and Greeks. We only have to
consider the things which are natural and necessary to all mankind. These are
open to all [to get] in the same way, and in [all] these there is no distinction
of barbarian or Greek. For we all breathe out into the air by the mouth and
the nose, and we [all eat with our hands]… (DK B44; Sprague 2001, p. 220;
emphasis added).
Lest we think there is present a classist distinction between “good”

families (the aristocracy, the rulers?) and lesser social nuclei to mar the
startling novelty, for the age, of the statement implying “all men are
created equal”, a different version has:
We recognize and respect [the laws of nearby communities], whereas those
of communities far away we neither respect nor revere. In this, however,
we have become barbarized towards one another, whereas, in fact, as far as
nature is concerned, we are all equally adapted to being either barbarians or
Greeks… (Dillon & Gergel 2003, p. 150; emphasis added).
This is what is meant by man the measure of all things. At a time

of growing Athenian pride and hegemony, stating that, “in theory”,
Greeks and barbarians might be alike, was not a good thing to be
teaching people, especially of the lower classes. The multifaceted
elaborations on the conduct of the orator first by Aristotle and then by
Cicero and Quintilian were still far in the future, so the early rhetori-
cians were possibly their own worst enemies because they did not
know what power they had unleashed. Yet what must be perceived here
as relevant to our inquiry is that there is a search for a new standard,
a frame of reference that does not discount the material conditions
or existence of the speakers as well as the social institutions wherein
human discourse occurs. And in order to forestall the tricky turning
of tables practiced by Socrates on his naïve interlocutors, which is
centered on the individual in relation to cosmos, we ought perhaps con-
sider the recent suggestion by one classicist who, having reconstructed

the actual usage of certain terms in the V Century, believes that what
is meant by the statement, Pantôn chrêmatôn metron estis anthrôpos,
(DK80 B1) is more along the following tenor: “Of everything and any-
thing the measure [truly-is] human[ity]…”5. It is mankind, humankind,
the first and last frame of reference, not God, not eternity, but mortals
in time, people in actual experience. God, eternity, and connexed val-
ues such as Unity and Perfection will have to be Translated into man-
ageable frames through discourse, that is, rhetorically, in history. This
should be kept in the background as we touch upon a few topics which
reflect the need, for philosophy, to rethink, in our post-metaphysical
and unpoetic times, the relevance of sophistic thought. For we are at a
stage where Theory, Discourse, and Method are not as clearly marked
as they will be after the Platonic deluge and the Aristotelean recon-
struction and classification of all things created.
For nearly twenty-three centuries the sophists have been associated,
and negatively, with the introduction on the cultural scene of Greece
of rhetoric as a special discipline. Why negatively is still a matter for
debate, but some of the reasons can be gleaned by reviewing the reac-
tions to them in their own time and the first generations thereafter. The
sophists had raised awareness of the import of language as language
in the determination of the relation mind-world, humans-nature, as we
will discuss in a moment, yet the word rhetoric actually did not even
exist when Protagoras, Prodicus, Gorgias and Antiphon were admir-
ing the work of Ictinus and Phidias on the Acropolis. Rhétorikê is a
fourth century coin which ought not to be imposed on fifth century
intellectuals, and on Corax and Lysias as the “founders” of the disci-
pline. The sophists were rather more concerned with the elusive logos,
5
Schiappa 1991, p. 121. Besides an acute reconstruction of the lexico-semantic
and stylistic distribution of key terms, Schiappa makes a strong case for a Protagoras
who responds to, rejecting it, the extremism of Parmenides, freeing the interpretive
horizon to accept and develop the Heraclitean perspective. For a different translation,
which I cannot comment on here but is implicitly relevant, see David K. Glidden,
where we read “Man is the measure of all states of affairs, of what is the case, that it
is the case, of what is not the case, that it is not so” (my emphasis).
68 PETER CARRAVETTA
specifically with its proper usage, orthos logos, in part to distinguish

themselves from the Eleatics6. Beyond that, they were interested in
democratizing the arête, a term which meant excellence until Plato
altered or reduced its sense to reflect the more abstract notion of
Virtue. Justice, another key word in the Platonic corpus, at the time of
the sophists was not dikê, but dikaiosunê, which also meant “personal
excellence”. Following Havelock both Kerferd and Schiappa empha-
size how the sophists actually addressed the issue of conduct in politi-
cal discourse, given that popular courts were a new feature of social
life and required hitherto unknown protocols of governing debate, or
opposing arguments. The horizon of interpretation therefore includes
something more, or other, than the even earlier paradigm shift away
from the “poetic” and the “mythological” mind set. We should rather
read the sophists as proleptically planting the seeds of what later would
be called the anthropology of religion and the rhetoric of argumenta-
tion7. More than that, we should read them as philosophers plain and
simple.
But in order to accomplish that, we must move from the hermeneusis
of To Say toward what was now needed, a configuration of To Explain,
which points precisely to the emergence of a self-consciousness about
how to use speech, how to make an other understand and yes, convince
him or her that one’s position is the best among other contending ones.
The misguided debate about and condemnation of the double-argument
in Protagoras and in the Dissoi Logoi8, was meant to discredit the soph-
6
Cf Schiappa 1991, pp. 40 sgg. The author argues, convincingly, that the word
rhetoric may have been coined by Plato when he wrote his Gorgias, in 385 BC, on the
model of other Platonic lexical creations ending in –ikê, such as eristikê, dialektikê,
antilogikê: “it would be remarkable if rhétorikê was not invented by Plato” (44). The
rhêtôr on the other hand was long known as “a politician who put forth motions in
court or the assembly” (ib.).
7
Some of the philological notes here resonate with the work of Chaim
Perelman.
8
Cf. in Dillon & Gergel 2003, pp. 318-333 and Diels-Kranz in Sprague 2001, pp.
279-293 for this Doric document, probably a set of student notes, made up of argu-
ments and counterguments, which has been associated with Protagoras’ two-logoi
thematic, and analyses by both Kerferd 1999 and Schiappa 1991, pp. 89-102.
ists because one of their claims was that of every argument there can
be a counterargument. Here we must invoke hermeneutics even in the
sense of To Translate in order to clarify how two different explanations
work and impact on the community. The older translations have: “On
every issue there are two arguments opposed to each other” (Michael
O’Brian); or “On every question there are two speeches [logoi], which
stand in opposition to one another” (Theodor Gomperz); or “there are
two opposite arguments on every subject” (Guthrie). These have been
called subjectivist translations and tend to make Protagoras merely a
rhetorician (in the negative sense) advocating that debate is possible on
any topic and anyone can contradict someone else. But a careful recon-
struction of the actual usage of the two key words in the aphorism,
namely logos and pragmata, would not give us “issue”, “question” or
“subject”, terms which our speaking and conceptualizing habits of the
past two or three centuries find appropriate, but which reflect a mental-
ity that has accepted logical dichotomies and the subject-object opposi-
tion. Rather, by what Schiappa calls a Heraclitean interpretation of the
passage, the word pragmata ought to be rendered with “things”, as in
the locution “hand me that thing”, which is an object, or “it seemed the
thing to do”, which is a deed or act (Schiappa 1991). Pragma meant
“reality” to Protagoras, and Untersteiner appropriately keeps this mind
when he translates: “In every experience there are two logoi in opposi-
tion to each other” (Untersteiner 1954, p. 19; my emphasis). The point
is that Protagoras was talking about the world, something outside of
the control of the speaker, the implication now being that whenever we
speak about how we relate to reality, there are at least two different
and implicitly opposing ways to account for or explain the experience
(Hence the second part of the “man is the measure” citation above
intercalated by Plato in his Protagoras). However, with the new ren-
dition we can fairly attempt to attribute to the historical Protagoras
the development and legitimate recasting of the Heraclitean theory of
flux and the presumed unity of opposites, insofar as both are present
whenever we assess “what is”. An even more crucial question at this
juncture is that in the attempt to understand the concrete world of
interpersonal relations, there is a novel dimension or space opened up
which makes the speaker aware that a way of mastering the world must
70 PETER CARRAVETTA
be devised, a preoccupation which in turn forces a distinction between

the natural world and the world of culture. And this, which I consider
the greatest contribution of the sophists in general and Protagoras in
particular, can be done only through the logos, or better, and more real-
istically, by means of diverse logoi, that is, the plurality of possibilities
afforded by language as used by a variety of speakers, an awareness
which seemed to offer previously unthought possibilities9. And, inevi-
tably, this will spill over into the question of knowledge, for Protagoras
believed knowledge is possible for man.
Here we are confronted with new interpretive horizons. We learn
that for the first time, knowledge is grounded in what can be estab-
lished to be the case (Untersteiner 1954, pp. 53 sgg.), which requires
precisely that two positions be weighted and compared, though owing
to post-Aristotelean habits of framing questions, we seem to prefer the
agonistic language of opposition and conflict. In essence we can see in
Protagoras the telescoping of three distinct moments originally derived
from presocratic reflection but developed in novel ways, which turn on
our adopted metacritical term of hermeneutics as To Explain. Adapting
to our need an assessment by Julius Moravcsik, we have:
a. explanation in terms of origin,
b. explanation in terms of constituency or “stuff,” and
c. explanation in terms of entities and attributes10.
We are not yet up to the theorization of the distinction between
substance and qualities, which will be taken up in Plato’s dialogue
Protagoras. Readers may recall the query, referred to above: is the
wind that feels hot to me and cold to you the same wind? Are hot and
cold properties of the wind, or are they possible effects which surge
up when a sentient being experiences the wind? In Protagoras’ extant
writings this conundrum has traditionally been understood reductively
9
This will make Plato’s Cratylus seem irretrievably old-fashioned as soon as it
was conceived.
10
Cited in Schiappa 1991, p. 95. Moravcsik’s article, “Heraclitean Concepts and
Explanations,” appeared in the anthology Language and Thought in Early Greek
Philosophy, ed. K. Robb, 1983.
as suggesting some sort of individualist anarchy or an atomization of

perspectives, yet what is being introduced here is a third element which
speaks directly to the retranslation of “man is the measure” as, in my
own words, “humanity is what constitutes the metron, the criterion, of
our understanding”: the wind may feel hot to you and cold to me, but
if the temperature is 100 degrees Fahrenheit there is a probability that
we both will state that this is a good thing, whereas if we see a house
on fire down the block we may both observe and agree that wind at
this moment is not a good thing. What lurks underneath is the future
problem of the relationship between meaning and reference.
In the second of the three stages of To Explain just mentioned,
which concerns constituency or the quid of what is out there, mean-
ing becomes focal in the “objective” or external recognition that there
exists a doxa which turns out to confer certain (semantic, symbolic)
values to language-use. It is in the sphere of this now necessarily
“included third” element, and through it, that we can begin to speak of
what is common knowledge, the basis of sensus communis, so we may
communicate at all, relying on the embedded encyclopedia of a par-
ticular society. Reference will turn out to be a factor in establishing the
validity of factual or uncontestable statements, and this will eventually
split into separate formal scales of values extending between concrete
and abstract, the empirical and the propositional. But before we elabo-
rate on this, let us address two more topics which will allow us to set
up the context for an assessment of where Discourse – not yet rhetoric!
– stands between Method and Theory, and what kind of hermeneutics
was possible in pre-Platonic times which may still be relevant to us.
These are the Stronger versus Weaker argument, and the unnerving
possibility of an entrenched relativism in human affairs.
Since the fourth century BC, the sophists have really been a thorn
on the side of philosophy, science, history and literary studies11. Maybe
11
Writes Schiappa: “Protagoras has been called the first positivist, the first
humanist, the forerunner of pragmatism, a skeptic, an existentialist, a phenomenalist,
an empiricist, an early utilitarian, a subjective relativist, and an objective relativist.
(15) He furnishes a detailed bibliography of all the scholars who wrote proving the
validity of each of these attributive labels (19n), though in the balance it appears the
72 PETER CARRAVETTA
it is time to live with the discomfort and abandon all hope of ever
coming up with tidy one-two answers to all problems, as the American
mass media persistently tries to do. Reality is more complex and our
addiction to categories and disciplinary boundaries often makes it
difficult to comprehend how a thinker thinks and speaks and whose
understanding require we mentally straddle several camps at once. The
sophists made discourse the center of their teaching, more concerned
with logos (but, again, in an entirely different ways than the Eleatics)
than with rhetoric (because it was a later attribution), and Protagoras
has been called the first to develop what later became known the
Socratic method in education. This polyvalent yet structured approach
required the invention of a metadiscourse, the earliest grammar:
Protagoras was said to have been the first to divide up discourse (logos),
according to one account into wish, question, answer and command, accord-
ing to another into narration, question, answer, command, reported narrative,
wish and summons, while the sophist Alcidamas proposed a different, four-
fold, classification, into assertion, negation, question and address (DK 80A1,
paragraphs 53-54). In addition Protagoras distinguished the three genders of
names, as masculine, feminine and those referring to inanimate objects (DK
80A27; Kerferd 1999, p. 68).
It follows then that beyond correct diction (orthoepeia) and correct-
ness of names (orthotés onomatón) what had to be taught ended up being
a process of determining what words were appropriate for a specific
situation, what something so named is in context, as opposed to some-
thing else bearing a different name, thus raising simultaneously issues of
morphology, semantics and reference12. Two developments ensue at this
juncture. One is that, on the way to rhetorical formalization, four dif-
ferent types of discourses can be identified that coordinate and circum-
scribe a discussion, one of which include a philosophical research. These
greatest number can be grouped in two hybrid categories: Protagoras as a humanist

and pragmaticist ante litteram, and Protagoras as an objective relativist and subjec-
tive relativist, these latter two related in a chiastic manner.
12
On the relevance of Protagoras as first grammarian see Glen Most, who sees
in the emphasis on the new relation set up between name and thing the shaping of a
protohermeneutics.
are eristic, antilogic, dialectic and elenchus. By the time we get to the
Platonic dialogues, they are pretty much stabilized, but at the time of the
historical Socrates and before, their range and deployment still overlap.
We have mentioned briefly the Two-logoi Argument. The accusa-
tion against the sophists was that in this fashion we could never say
anything about anything, whereas what I think was being thematized
in a yet hypothetical way is that language is one thing, and the phenom-
enal world another, and that it is reasonable, and possible, that about the
same entity two different or contrasting things can be said. In fact this
condition bothers only those, mainly Platonists (known in postmod-
ern critique as essentialists) and religious types, who must ground the
meaning of a sentence in some supratemporal Axiom or Form or God.
Furthermore, the power of a logos thus unrestrained means that, the
question of a Truth with a capital letter losing its purchase on human-
ity, anyone can make a case if he or she argues appropriately, that is,
persuasively, with force, relying on the perceptible perennial mobil-
ity of things. This introduces, as we have been saying all along, the
cruciality of place, which we can describe as the temporally marked
co-occurrence, and appropriate coalescence, of speech and situation of
the existent. The relativism which was so abhorrent to many who pre-
ferred single-answer, authoritarian solutions is in fact necessary to any
conception of dialogue if we think of speakers in given real-world situ-
ation, or what we have called above Contexts and Horizons. This posi-
tion is valid if we wish to continue to believe that at the time of Pericles
there was indeed a youthful democracy in the offing in which, as in all
democracies, everyone has a right to defend his or her own beliefs and,
moreover, and most importantly where and when that is impossible
for some reason, someone else can stand up for them. What Plato and
the great systematizer Aristotle did was consider the technique of anti-
logic argument solely on the logical plane, not on existential or political
grounds. The reason for this power move is obvious: anti-logic demon-
strates that there is no innate coherence in phenomena, and that what is
called the real is itself illogical: what the idealists and the realists of a
succeeding generation cannot bear, the sophists accept and deal with.
The sophists thus are faced with, and attempt to explain, a character-
istic of language in action which predates or subtends heuristics, and that
74 PETER CARRAVETTA
is eristics, which is fundamentally an attitude or a desire to make one’s

reasons, or beliefs, stand against the inevitable counterargument13. In
his translation of Protagoras’ dictum, as reported by Aristotle (Rhetoric
1402a23), Lane Cooper asserts that what is detestable about the soph-
ists is their penchant for “making the worse appear the better cause”.
This reductive, “perverse version” (Schiappa 1991, p. 103), reflects the
tone of scores of other indictments of rhetoric which lent credence to
the myth that “rhetorical speech” – as if there were any other kind! – is
not concerned with the truth, the sophists as a school being no less than
profiteers and so on14. The fact is that what is intended is that the rhetor
will essay to make “one argument stronger than another” (Schiappa
1991, pp. 107-111; my emphasis), which is perfectly plausible and realis-
tic in real-life situations, it having precious little to do with the concept
of Truth. This of course is linked to another of Protagoras’ problematic
aphorisms, namely, that “it is impossible to contradict”15. Clearly one
cannot say that A is B and A is not B at the same time. What is left out
of a more correct evaluation are connexed markers or marginalia which
always go with discourse, insofar as discourse is never pure, not even
13
Cf. on this Paolo Valesio: “every discourse in its functional aspect is based
on a relatively limited set of mechanisms…that reduce every referential choice to
a formal choice”; and since “it is never a question – or at least, never primarily and
directly a question – of pointing to referents in the ‘real’ world, of distinguishing
true from false, right from wrong, beautiful from ugly, and so forth”, the choice is
“only between what mechanisms to employ, and these mechanisms already condition
every discourse since they are simplified representations of reality, inevitably and
intrinsically slanted in a aprtisan direction”. Thus these mechanisms may appear
“to be gnoseological, but in reality they are eristic: they give a positive or a negative
connotation to the image of an entity they describe in the very moment in which they
start describing”. (Novantiqua, 21-2; emphasis in the original)
14
These include scholars of the rank of Keith Erickson, Alexander Sesonske, and
W. K. C. Guthrie. See in particular Sesonske, “To make the weaker argument defeat
the Stronger” in Erickson 1979, pp. 71-90 for an expression of contempt for the soph-
ists no longer acceptable by the scholarship of the past twenty years.
15
There is no passage that can be attributed to Protagoras with certainty which
says ouk estin antilegein, it is always reported by secondary sources, from Plato
through Diogenes Laertius. Taken veridically as representative of a logical statement,
Aristotle can quite easily dismiss it (cf. Metaphysics 1005b19-20 and 1024b34).
when it seeks, and believes it has achieved, the crystalline transparency

of the ideal metalanguage (symbolic logic, mathematics, the periodic
table, etc.). These elements, which clearly belong in our interpretive
model, are: time, situation, context. For it is perfectly coherent, indeed
“logical”, to assert that A is B at Time t1, and A is not B at Time t2.
There is no antilogic understood in the post-Aristotelean sense, nor
unethical implication in Protagoras16, just a pedagogic aim which sal-
vages the background consideration that one argument, A, which wins
over another, B, albeit about the same object, C, is still allowing the
givenness and relevance of C to exist and be given a significant role in a
different time and place, as it attributes relevance in the construction of
meaning to something outside of the speakers, something undetermined
by either A or B. It is Discourse which is allowing at the same time an
exchange, even a confrontation, between two possible speakers without
assuming that whoever wins or loses the argument is taking the world
down with them. In other words, it is not, as it will be in Plato, a ques-
tion that one argument, A, cannot co-exist with another if the speaker
states that it is a negation, not-A, or if he asserts that A is B – think of
the way Socrates leads Theaetetus to agree that perception = appear-
ance = being! – but, rather, that in the exchange A will convince more
people to believe in its understanding, or accept its own version, of a
given episteme – and given a whole set of other factors that always
surround the mise-en-scene of language, such as character, mood, time,
place, audience, etc. – than perhaps B is able to do. One is not better
– logically, ethically – than the other: one just carries the day. That is
why we have assemblies. In a democracy, laws can be promulgated and
ratified, but also revoked or abrogated. Those who violate them, have a
chance to explain why, as Antiphon’s Tetralogy amply documents.
It is this phenomenalism which, despite being rooted in a common-
sensical empirism, makes critics incapable of seeing the philosophical-
ly positive aspects of Protagoras’ relativism. Phenomena and Discourse
16
Dumont makes a case that in presocratic philosophy the discussion on moral-
ity is quite limited, and we shouldn’t retroject it onto the sophists. Similar position is
implied in Olfray.
76 PETER CARRAVETTA
inhabit two altogether different spheres, but are related in lived time.
Phenomena occur and humans try to make sense of them. They recur
to logos for that. But though logos is still an undifferentiated expres-
sive plenum, yet it can be understood as capturing at least three differ-
ent aspects of phenomena. Following but modifying Kerferd, we can
therefore distinguish three traits for possible real-world application of a
logos deprived, by the V center of its mythico-mystical aura:
1. Metalinguistic: logos as grammar-cum-rhetoric, which includes, as we
have seen, statements, arguments, descriptions and other metaterms;
2. Logic: mental processes such as thinking, reasoning, accounting and
explaining;
3. Reference: the world as offering context and horizon, the dreaded vari-
able referent out there, that about which we are able to speak, which
includes abstract constructs such as principles, formulas, natural laws
insofar as they are “regarded as actually present in and exhibited in the
world-process” (Kerferd 1999, p. 83).
Understood that in any given speech act all three aspects may and
usually do co-exist, we must nonetheless attempt to isolate and posit a
few corollaries:
a. the very question of Being is either overcome or made irrelevant
except as a topic or theme just like any other and subject to ideological
scrutiny;
b. the question of reality acquires a more compelling and for the moment
ambiguous aspect;
c. and the position of the speaker which becomes that of examiner and
judge, invested with the task of determining what features of the real,
of the surrounding phenomena, can be said to bear such and such a
meaning even while they may exist factually or may not de facto exist
at all but can legitimately be supposed to exit, “for the sake of argu-
ment”, such as unicorns and utopias.
We shall ignore Gorgia’s radical nihilism, but will make room for
Antiphon’s psychological approach, which veers toward an elementary
form of subjectivism. There is no “solution” nor “answer” to the con-
ceptual fields thus opened up. The possible resolutions will occupy
Plato and Aristotle their entire lives, the first by developing both, meth-
od and theory (where the latter term is understood as a superior unitary
frame of reference on the basis of which the distinction between the

truth of a statement and absolute truth can be made); while Aristotle
worked on the notions of quality and predication.
In Protagoras’ world the one thing we can be certain of is that there
is a language that permits exchange and reveals effects on the organi-
zation of the social world, on that nous which has now left concerns
for the phusis to the next generation of medics, astronomers, natural
scientists, philosophers. What Protagoras, Prodicus, Thrasymachus,
and Antiphon did understand is the relativity of positions in the social
spectrum, the uncomfortable truism that might makes right and that
therefore institutions and education deserved attention, that eristic
seems to haunt heuristic constructs, that virtue can be taught, that
change, becoming, is an essential element in language and human
understanding, that the probable ought to be taken into consideration
in any evaluation of a course of action, including metaphysical and
epistemological decisions, and finally that interpretation is forever
swimming in a sea of possibilities.
From the standpoint of our initial premises17, Theory and Method
are both negotiable in terms of what value we ascribe to Discourse. In
terms of our model, Individual and Society are clearly ab initio related
and co-enabling, the Work lends itself to varying meanings on the
basis of time, place and context, the responsibility of the Interpreter
is less theological or metaphysically biased and more subject to the
eventualities of social and political intercourse. But the criterion for
interprertation now rests solely with homo humanus.
17
This last paragraph refers to the hermeneutical model developed in my
forthcoming book The Elusive Hermes, from which this article is excerpted. The
model stresses the pragmatic dimension of the rhetorical underpinnings of both
theory (as vision) and method (as application). Its basic elements are constituted
by the necessary co-presence, when interpreting something, of an Interpreter, a
Society, the Work in question, and a Language act (whether metalinguistic or con-
versational). As I have corrected galleys during the Summer of 2011, I can furnish
the full bibliographic information: Peter Carravetta, The Elusive Hermes. Method,
Discourse, and the Critique of Interpretation, Aurora (Colorado), The Davies
Group Publishers, 2011.
78 PETER CARRAVETTA
Bibliography.
Arduini 2004 Stefano Arduini, La ragione retorica, Rimini, Guaraldi.
Aristotle 1984 Aristotle, The Complete Works, 2 vols., ed. Jonathan
Barnes, Princeton, Princeton UP.
Cassin 1986 Positions de la Sophistique, ed. Barbara Cassin, Paris,
Vrin.
Cherwitz 1990 Rhetoric and Philosophy, ed. Richard A. Cherwitz,
London, Lawrence Erlbaum.
Consigny 1996 Scott Consigny, Edward Schiappa’s Reading of the
Sophists, in «Rhetoric Review», 14, No. 2, pp. 253-269.
Cornford 1986 Francis Macdonald Cornford, Before and After Socrates,
London, Cambridge UP, [1932].
Dillon & Gergel 2003 The Greek Sophists, eds. John Dillon and Tania Gergel,
New York, Penguin.
Dumont 1986 Jean-Paul Dumont, Prodicos: de la méthode au système,
in Cassin 1986, pp. 221-232.
Enos 1976 Richard L. Enos, The Epistemology of Gorgia’s
Rhetoric: A Re-examination, in «Southern Speech
Communication», no. 42, pp. 35-51.
Erickson 1974 Aristotle: The Classical Heritage of Rhetoric, ed. Keith
V. Erickson, Metuchen (NJ), The Scarecrow Press
Erickson 1979 Plato: True and Sophistic Rhetoric, ed. Keith V. Erickson,
Amsterdam, Rodopi
Glidden 1975 David K. Glidden, Protagorean Relativism and Physis,
in «Phronesis», vol. 20, No. 3, pp. 209-227.
Guthrie 1962 W. K. C. Guthrie, A History of Greek Philosophy, I,
Cambridge, Cambridge UP.
Hamilton & Cairns Plato, The Collected Dialogues, eds. E. Hamilton & H.
1978 Cairns, Princeton, Princeton UP.
Hodges 1996 Karen Hodges, Unfolding Sophistic and Humanistic
Practice through Ingenium, in «Rhetoric Review», vol.
15, No. 1 (Autumn), pp. 86-92.
Kerferd 1981 The Sophistic Movement. Cambridge, Cambridge UP.
Kerferd 1999 G. B. Kerferd, 2003. “The Sophists,” in C.C.W. Taylor
244-70. 1999 [1980].
Kirby 1990 John T. Kirby, The ‘Great Triangle’ in Early Greek

Rhetoric and Poetics, in «Rhetorica», VIII, 3 (Summer),
pp. 213-228.
Kirk & Raven & The Presocratic Philosophers, eds. G. S. Kirk, J. E.
Schofield 1983 Raven, M. Schofield, 2nd ed., Cambridge, Cambridge
UP, [1957].
McCoy 2008 Marina Mc Coy, Plato on the Rhetoric of Philosophers
and Sophists, Cambridge, Cambridge UP.
Manetti 1993 Giovanni Manetti, Theories of the Sign in Classical
Antiquity, Eng. Transl. by C. Richardson, Bloomington,
Indiana UP.
Most 1986 W. Glenn, Sophistique et herméneutique, in Cassin
1986, pp. 233-245.
Nestle 1942 Wilhelm Nestle, Vom Mythos zum Logos, Stuttgart,
Kroner.
Nestle 1968 Wilhelm Nestle, Die Nachsokratiker, Düsseldorf,
Scientia, [1923].
Onfray 2006 Michel Onfray, Le Sagesse antiques. Contre-histoire de
la philosophie, I, Paris, Grasset.
Payne 1986 David Payne, Rhetoric Reality and Knowledge: A
Re-examination of Protagoras’ Concept of Rhetoric, in
«Rhetoric Society Quarterly», 16, pp. 187-197.
Poulakos 1995 John Poulakos, Sophistical Rhetoric and Its Early
Reception, Columbia, South Columbia UP.
Pullman 1994 George L. Pullman, Reconsidering Sophistic Rhetoric in
Light of Skeptical Epistemology, in «Rhetoric Review»,
vol. 13, No. 1 (Autumn), pp. 50-68.
Quimby 1974 Rollin W. Quimby, The Growth of Plato’s peerception of
Rhetoric, in «Philosophy and Rhetoric», 7, pp. 71-79.
Robb 1990 Language and Thought in Early Greek Philosophy, ed.
Robb Kevin, Philadelphia, Pennsylvania UP.
Robinson 1979 Contrasting Arguments: an edition of the Dissoi logoi,
ed. T. M. Robinson, New York, Arno P.
Roochnik 1995 David Roochnik, Socratic Ignorance as Complex Irony:
A Critique of Gregory Vlastos, in «Arethusa», 28, pp.
39-52.
80 PETER CARRAVETTA
Ruggiu 1975. Luigi Ruggiu, Parmenide, Venezia, Marsilio.

Scenters-Zapico 1993. John Scenters-Zapico, The Case for the Sophists, in
«Rhetoric Review», vol. 11, No. 2 (Spring), pp. 352-367.
Schiappa 1991 Edward Schiappa, Protagoras and Logos: a Study
in Greek Philosophy and Rhetoric, Columbia, South
Carolina UP.
Schmidt 1995 The Specter of Relativism. Truth, Dialogue, and
Phronesis in Philosophical Hermeneutics, ed. Lawrence
K. Schmidt, Evanston, Northwestern UP.
Scott 2002 Does Socrates Have a Method?, ed. Gary A. Scottt,
University Park, Pennsylvania State UP.
Sprague 2001 The Older Sophists, ed. Rosamond K. Sprague,
Indianapolis, Hackett Publishing.
Taylor 2001 From the Beginning to Plato, ed. C. C. W. Taylor, New
York, Routledge.
Untersteiner 1954 Mario Untersteiner, The Sophists, Engl. Transl. by K.
Freeman, Oxford, Blackwell.
Valesio 1980 Paolo Valesio, Novantiqua. Rhetorics as a Contemporary
Theory, Bloomington, Indiana UP.
Valesio 1986 Paolo Valesio, Ascoltare il silenzio. La retorica come
teoria, Bologna, Il Mulino.
Viano 1965 Carlo Augusto Viano, Retorica, magia e natura in
Platone, in «Rivista di filosofia», LVI, 4, ottobre-
dicembre.
Vernant 1983 Jean-Pierre Vernant, Myth and Thought among the
Greeks, London, Routledge & Kegan Paul, [1965].
Zeller 1951 Eduard Zeller, Outlines of the History of Greek
Philosophy, Engl. Transl. by L. R. Palmer, Bristol,
Thoemme P.
MICHELE PRANDI
FORMAL CONTRADICTION AND CONSISTENT THOUGHT:

OXYMORON*
For health and disease have for subject the

body of some living creature, and whiteness
and blackness a body which need not be
specified further. And justice, likewise, and
injustice arise in the souls of mankind.
Of contraries this, too, holds good, that,

the subject remaining identical, either may
change to the other […]. What is healthy
may well become sick, what is white may in
time become black, what is cold may in time
become hot. And the good becomes bad, the
bad good.
Aristotle
If we share the idea that living tropes are in the first place strategies
of textual and discursive interpretation of conflictual complex mean-
ings (Prandi 1992; 2004: Ch. 11; 2007), oxymoron can be defined a
figure based on a specific kind of conflict, that is, on contradiction.
Contradiction takes place when two opposite terms are syntactically
*
I am grateful to Derek Boothman and Fiachra Stockman for the revision of
English expression, and to Francesco Bertolini, Derek Boothman, Remo Bracchi,
Claudia Bussolino, Elisa Caligiana, Amedeo G. Conte, Pierluigi Cuzzolin, Hanna
Flieger, Francesco Giardinazzo, Patrick Leech, Giulio Prandi, Monica Savoca,
Elisabetta Spediacci, for plenty of examples and hints. Many thanks to Sara Piccioni
for translating Spanish and Italian examples, and to Paolo Rambelli for the transla-
tions from Latin.
82 MICHELE PRANDI
connected within a single expression uttered by a single speaker and

without temporal cleavage. The most transparent form of contradic-
tion takes place when two opposite predicates are applied to the same
subject: for instance, I both wished and feared to see Mr Rochester
(Charlotte Brontë).
The main idea that lies at the ground of this paper can be put in a
nutshell. Contradiction is a formal kind of conflict, which challenges
the formal organisation of concepts into oppositions independently of
their substantive content. Inconsistency is a substantive kind of con-
flict, which challenges the organisation of conceptual contents into
consistent relational networks. As a formal kind of conflict, contradic-
tion is compatible with conceptual consistency. Whereas synecdoche,
metonymy and metaphor typically stem from inconsistent complex
meanings, oxymoron combines a formal contradiction with a consist-
ent conceptual content. In order to illustrate this point, let us compare a
typical instance of inconsistent meaning and a typical instance of con-
tradiction. The former is the first line of Alcman’s famous Nocturne1,
They sleep, the mountain peaks; the latter is Catullus’ well-known
fragment:
Odi et amo. Quare id faciam, fortasse requiris.
Nescio, sed fieri sentio et excrucior.
I hate and I love. You might ask why I do it.

I do not know, but I feel it happening and I am excruciated.
The inconsistent complex meaning They sleep, the mountain peaks
attributes an attitude restricted to animate beings to an inanimate sub-
ject, calling into question its conceptual identity. This conflict is ready
to be interpreted as a metonymy, if sleep is attributed to some animate
being connected with mountains through a consistent relation: sleep-
ing animate beings live in mountains. But it is also ready to be taken
as a metaphor, if either sleep is projected upon a consistent state of the
mountains – silence, for instance, is seen as a kind of sleeping – or the
consistent subjects of sleeping are projected upon mountains, treating
1
West 1993, p. 35.
FORMAL CONTRADICTION AND CONSISTENT THOUGHT: OXYMORON 83
them as if they were animate beings. If the expression is interpreted as

a metonymy, the conflict is defused and consistency is restored. If it
is interpreted as a metaphor, the conflict is valorised and turned into
an instrument of conceptual creation. Owing to this, only metaphor
can be considered a true figure of inconsistency, and therefore the real
counterpart of oxymoron, the figure of contradiction.
The contradiction Odi et amo attributes a human attitude to a
human being. The conflict does not involve the conceptual identity of
the subject, but the formal organisation of concepts into oppositions.
As opposite concepts, hate and love cannot be predicated of the same
subject without contradiction: a man is supposed either to hate or to
love the same woman2. In spite of the formal contradiction, however, it
is a fact that both love and hate can be attributed to a human being in
a consistent way. This complex conceptual structure, which associates
formal contradiction and conceptual consistency, is what justifies the
strategies of textual use and interpretation open to oxymoron.
The example analysed above certainly shows that it is possible
for a formal contradiction to depict a consistent state of affairs, and
therefore a consistent thought. However, this property cannot be gener-
alised. The instance of contradiction we have just examined combines
a set of formal and conceptual properties that cannot be taken for
granted for any kind of contradictory expression. The formal syntactic
structure of the expression applies two opposite predicates to a subject.
In this way, the predicative relation is not directly involved into the
formal contradiction, which is internal to the predicate. The subject
is an individual subsumed under a classificatory concept, whereas the
2
Of course, the utterance is contradictory only if the direct objects of the
opposite verbs - here latent - refer to the same referent; otherwise it is consistent:
Remember me, but ah! Forget my fate! (Nahum Tate). More generally, contradic-
tion depends on the identity of saturation. In such an expression as Every limit is
a beginning as well as an ending (G. Eliot), contradiction is only apparent, for the
complements of the opposite terms – beginning and end – refer to different circum-
stances: the end of the novel is the beginning of a new life for its characters. When
Rushdie writes Death, most present of absences, had entered the garden, and from
that moment on the absences multiplied, the presence of death entails the absence of
many living beings.
84 MICHELE PRANDI
predicate opposes two processes, that is, two relational concepts. Now,
we can imagine a wide range of expressions that combine in different
ways both different formal frames and different kinds of concept, with
very different outcomes.
First, we can imagine different syntactic distributions of the correl-
ative terms. Beside co-occurring within the predicate, the correlative
concepts can directly form a predicative link – This cottage is not a cot-
tage – and even co-occur within the borders of a noun phrase: His sad
joy. Second, different distributions of different kinds of concept can fill
each of these formal structures. In His joy is sad and His sad joy, for
instance, both subject and predicate, head and modifier are relational
concepts. In This cottage is not a cottage and This cottage is a castle,
both subject and predicate are classificatory concepts. The opposite
concepts, in turn, can be either the members of a lexical paradigm –
Mary is happy and sad – or a term and its negation: Mary is and is not
sad. Thanks to the interplay of all these formal and conceptual factors,
some of these expressions end in contradiction – for instance His joy
is sad – and some do not, as for instance This cottage is a castle. In
the presence of different kinds of concepts, each form of contradiction
has its own way of conveying a consistent content. Mary is happy and
sad has not the same kind of content as Happiness is not happiness;
This cottage is old and new is very different from This cottage is not
a cottage. Owing to the rich typology of contradictory expressions and
to the variety of concepts and conceptual structures involved in them,
the correlation between formal contradiction and conceptual consist-
ency cannot be taken for granted but has to be called into question, and
checked against a wide range of different instances documented in real
texts. Two points in particular have to be examined: on what formal
and conceptual conditions a contradiction takes place, and whether, to
what extent, and to what formal and conceptual conditions a contradic-
tion is ready to convey a consistent conceptual content.
In the following pages, we shall try to answer all these interconnect-
ed questions. The latter point, that is, the compatibility between formal
contradiction and conceptual consistency, will be examined first and in
its general terms, taking into account clear cases of contradiction of the
most canonical type, that is, instances that jointly apply two opposite
relational predicates to an individual subject (§ 1). The former, that is,

on what formal and conceptual conditions a contradiction takes place,
requires a careful identification of both the forms open to the syntactic
framing of contradiction, and the forms and contents displayed by its
paradigmatic condition, that is, opposition. We shall carefully examine
the different forms of expression open to contradiction (§ 2), the dif-
ferent kinds of concepts, correlations and oppositions (§ 3) and their
distribution (§ 4), in order to explore what kinds of structure are ready
to convey a contradiction, what kinds of contradictory expressions are
ready to convey a consistent conceptual content, on what conditions
and with what outcomes.
1. Contradiction and consistency.
Die Wahrheit der Tautologie ist gewiß,

des Satzes möglich, der Kontradiktion
unmöglich
L. Wittgenstein
Contradiction is based on opposition, and like opposition is a for-

mal relation, independent of the conceptual contents involved. If P and
non-P are opposite concepts, a structure of the form x is P and non-P
frames a contradiction for whatever value of P3. The formal nature of
contradiction naturally begs a question, that is, to what extent the pres-
ence of a formal contradiction within an utterance affects the consist-
3
The symbolism is inspired by Wittgenstein (1929, p. 162): “We get the picture
of the pure form [of a proposition] if we abstract from the meaning of the single
words […] That is to say, if we substitute variables for the constants of the proposi-
tion”. This symbolism, however, has an intrinsic limit: it captures the formal nature
of oppositions and contradictions based on negation – for instance Mary is good and
non-good – but fails to grasp the formal nature of oppositions and contradictions
based on lexical forms – for instance Mary is good and bad – which are rooted in the
meaning of words (see § 1.4). When discussing the contradiction of the form x is P
and non-P, it will be useful to use non-P as referring not only to the negation of P,
but also, in a broad sense, to the opposite of P. If P means good, for instance, non-P
covers both non-good and bad.
86 MICHELE PRANDI
ency of its substantive conceptual content. This, in turn, leads us back

to a more general question, that is, to what extent the formal properties
of an expression affect its use, and, in our particular case, to what
extent the presence of a formal contradiction affects the coherence4 of
the connected speech act.
Within the logical tradition, it is normally assumed that the formal
logical properties of a contradiction are immediately transferred onto
both the conceptual content and the utterance act. This approach can be
traced back to Parmenides, who curses “mortals knowing nothing […]
by whom being and not-being have been thought both the same / And
not the same5”. More than two thousand years later, Strawson (1952, pp.
2-3) argues that uttering a contradiction amounts both to building up a
void conceptual content and to performing a contradictory, and therefore
self-defeating speech act: “a man who contradicts himself may have
succeeded in exercising his vocal chords. But from the point of view of
imparting information […] it is as if he had never opened his mouth. He
utters words, but does not say anything […] The point is that the standard
purpose of speech, the intention to communicate something, is frus-
trated by self-contradiction”. In fact, the relationship between the formal
properties of a contradictory expression, the conceptual properties of its
4
As stressed by Conte (1988, p. 29), textual coherence must be carefully dis-
tinguished from conceptual consistency. Consistency is the negative property of a
sentence in isolation, and links up with the absence of contradiction or conceptual
conflict in its meaning. Coherence is the positive property of the relation between an
utterance and the text or the communicative situation it is part of. The consistency of a
sentence meaning rests on a set of external criteria – on a true “grammar of concepts”
(Prandi 2004). The coherence of a text does not depend on a set of a priori external
requirements, but on purely internal criteria, which are as contingent as the textual
configuration itself. A text is not coherent because it fits some kind of grammar
independent of it, but because its parts fit one another, that is, they can be interpreted
as cooperating in attaining a unitary communicative goal. A small terminological
problem is connected with the use of the terms consistency and consistent, which are
commonly used to denote both the absence of formal contradiction and the absence
of substantial conceptual conflict. In this paper, the context will make clear in each
individual case which use is relevant.
5
Gallop 1984: Fragment 6.
content and the pragmatic properties of its textual and discursive use is
far from being so direct, and deserves careful examination.
It is true that uttering a contradiction amounts in the first place to
performing a contradictory speech act. It is also true that a speaker
can succeed in performing a coherent speech act while using a con-
tradiction, on condition that a cooperative addressee is ready to draw
a coherent message from the contradictory meaning thanks to a chain
of inferences. However, this implies neither that the incoherence of the
act depends on the presence of a formal contradiction in the meaning
of its linguistic instrument, nor that the conceptual consistency of a
contradictory meaning is in turn the outcome of an act of cooperative
inferencing on the part of the addressee. In fact, the incoherence of a
speech act is independent of the contradictory content of the involved
expression (§ 1.1), whereas the conceptual consistency of a contradic-
tion is not the outcome of special interpretative devices, but it is rooted
in the structural properties of the expression itself (§ 1.2).
1.1. Formal contradiction in speech acts.
Lis est in vocibus ipsis; Sed litem totam

sedat sententia vocum
G. de Vinsauf
Strawson’s remark about the consequences of uttering a contradic-

tion is certainly true in the first instance, but it could just as certainly
encourage a false inference, that is, the idea that the speech act is con-
tradictory because the expression involved is a contradiction.
Unlike a consistent utterance such as The book is on the table, a
contradiction cannot be used immediately to perform a consistent
speech act. Taken at its face value, a contradiction is necessarily false,
and it is owing to this that it is supposed not to be able to convey a
real message. If one looks deeper into the process of communication,
however, one realises that the use of a contradictory utterance is neither
a necessary nor a sufficient condition for performing a contradictory
speech act. On the one hand, the discourse behaviour of contradiction,
which is necessarily false, is shared by tautology, which is necessarily
88 MICHELE PRANDI
true. On the other hand, owing to the cleavage between the meaning
of the expression and the message it is meant to convey on a particular
occasion, the use of a contradiction is no obstacle towards performing
a consistent speech act.
There is certainly something in the meaning of a contradiction
that prevents it from being immediately taken as a coherent and
relevant message. This something, however, is not the presence of
a contradiction, but a more general property of which contradiction
is only a particular case. The proof is the behaviour of tautology in
discourse. Logically speaking, tautology is the opposite of contradic-
tion: whereas contradiction is both conflictual and necessarily false,
tautology is by definition consistent and necessarily true. In spite of
this, uttering a tautology is as contradictory a speech act as uttering
a contradiction. For opposite reasons, the speaker who utters a tautol-
ogy does not fulfil his communicative commitment. After engaging
himself in saying something, he says nothing. As Hegel (1812-1813: p.
415) points out “If, for example, to the question ‘What is a plant’ the
answer is given ‘A plant is - a plant’ […] we see that the beginning,
‘The plant is -’, sets out to say something, to bring forward a further
determination. But since only the same thing is repeated, the opposite
has happened, nothing has emerged. Such identical talk therefore
contradicts itself ”6.
The joint observation of contradiction and tautology as instruments
of speech acts shows that coherence and contradiction in speech acts
depend on the behaviour of the speaker rather than immediately on the
content of the expressions. A speaker who opens his mouth makes a
promise. If he does not fulfil it, he contradicts himself. Now, there are
6
Hegel’s remark is sound as far as the discursive use of tautology is concerned
– but the same criticism holds, as we have seen, for his favourite logical structure,
that is, contradiction. Hegel however, goes far beyond this point. The discursive value
of tautologies is taken as proof against the identity principle as leading principle of
consistent thought and speech: “Identity is contradictory”. When making this move,
Hegel clearly shares the same presupposition he attributes to the tradition he claims
WRFKDOOHQJHíQDPHO\DQLPPHGLDWHLGHQWLILFDWLRQEHWZHHQWKHORJLFDOSURSHUWLHVRI
the utterance and its behaviour in speech and thought.
many ways of not fulfilling the communicative contract. The use of

a contradictory expression is only one of them, along with the use of
a tautology. But also the use of a consistent synthetic predication like
The book is on the table can fail to fulfil the communicative contract,
if its content is not relevant, that is, if it is not coherent with the text or
the ongoing discourse. The difference is that such logically defective
expressions as tautology and contradiction fail for essential reasons,
rooted in their very structure, whereas a consistent synthetic predica-
tion fails for contingent reasons, rooted in its relationship with a con-
tingent communicative environment.
What has just been said holds for the meaning of the expression:
taken at its face value, it is not up to convey a coherent message. In
spite of this, a contradiction can succeed in performing a coherent
speech act on condition that its essential logical property is overturned
at the very moment of use, that is, on condition that a cooperative
addressee is able to infer from its conflicting meaning a consistent,
coherent and relevant message. In other words, on condition that the
contradiction principle, broken within the expression as a rule for
speaking, is applied by the cooperative interpreter as a rule for think-
ing. When Heracleitus writes that Into the same river we step and do
not step; we are and we are not7, for instance, the message is not so
difficult to grasp - it refers to the changeability of things and experi-
ences through irreversible time. This message, however, leads one far
away from the meaning of the expression: its content is not grounded
on the long-lasting structure of the expression, but it is the ephemeral
issue of a contingent act of interpretation performed by a cooperative
addressee. Just as contradiction, tautology can also be used to convey
a coherent and relevant message in actual texts or discourses: Poison is
poison. Tropical fever is tropical fever (Conrad); A river’s a river, and
if you’ve got a mill, you must have water to turn it (Eliot); Wickedness
is always wickedness, but folly is not always folly (Austen); An onion
is an onion is an onion (Hemingway). Once again, this requires that a
cooperative act of interpretation overturns the qualifying logical prop-
7
Jones 1959: Fragment LXXXI.
90 MICHELE PRANDI
erty of the expression, changing a void meaning into an informative

and relevant message8.
The cleavage between meaning and message displayed by both
contradiction and tautology is only a particular case of a more gener-
al property of verbal communication. The message a speaker intends
to communicate, which is the purpose of a speech act, is not the same
thing as the meaning of the expression he utters (Grice 1975; Sperber
& Wilson 1986). A meaning is the essential property of an expres-
sion, that is, a systematic and long-lasting structure of the symbolic
order. The occasional message it is entrusted with is a contingent
structure belonging to the indexical dimension: it is the content of
an individual’s intention and the aim of an individual’s action within
which the linguistic expression plays the role of instrument (Prandi
2004, Ch. 1). The two objects – a meaning and a message – happen
to coincide when an act of literal expression calls for an act of lit-
eral interpretation, but they are nonetheless heterogeneous kinds of
objects. If for some reason, due either to the expression itself, as in
the case of contradiction and tautology, or to the lack of coherence
with the communicative environment, the meaning of the expres-
sion is not adequate for being taken as the intended message, the
addressee of the speech act is ready to look for the relevant message
far away from it, thanks to a more or less complex chain of infer-
8
Wierzbicka (1987, p. 96) argues that most tautologies both display characteristic
and language-specific syntactic patterns and have highly conventional meanings:
“The constructions in question have a language-specific meaning, and this meaning
must be spelled out in appropriate semantic representations”. This fact is interpreted
by Wierzbicka as an argument against the idea that the interpretation of tautologies
is a creative textual or conversational act. Now, it is true that tautologies easily turn
into proverbs, the most typical kind of conventionalised meaning: for instance, Boys
are boys, A promise is a promise. However, conventionality and creative interpreta-
tion are not incompatible strategies, for both are within the horizon of any kind of
figurative speech, and in particular of metaphors. The tendency of tautology towards
conventionalisation is certainly not shared by contradiction, in spite of some exam-
ples such as To blew hot and cold (about an idea); The best form of defence is attack,
or Italian La miglior difesa è l’attacco; Si stava meglio quando si stava peggio (We
were better off when we were worse off ).
ences. In this case, an act of non-literal interpretation meets an act of

non-literal expression9.
The gap between meaning and message, and the contingent and
inferential nature of the latter, easily justify the aptitude of such logi-
cally defective expressions as tautology and contradiction to carry a
coherent message. More generally, the quality of an act of communica-
tion is measured by criteria internal to the action, that is, by its coher-
ence with its aims rather than by the logical properties of its instru-
ment. Insofar as a speaker succeeds in conveying the relevant message,
the logical properties of the instrument are a minor point.
1.2. The conceptual consistency of contradiction.
Lorsqu’une proposition n’est pas identique,

c’est à dire lorsque le prédicat n’est pas com-
pris expressément dans le sujet, il faut qu’il
y soit compris virtuellement
G. W. v. Leibniz
In the previous section, I have argued that the pragmatic properties

of a speech act are independent of the logical properties of the expres-
sion involved. In particular, the presence of a contradictory expression
is neither a necessary nor a sufficient condition for performing a con-
tradictory speech act. On the one hand, the use of non-contradictory
9
Such properties as literal and non-literal do not apply to the meaning of expres-
sions, but to the twofold relationship between a meaning and a message, that is, to
both expression and interpretation (see Prandi 2007). An intended message can either
coincide with the meaning of the expression used to convey it or go far beyond. In
the former case, the expression is literal; in the latter, it is non-literal. Conversely,
an act of interpretation can either take the meaning of the utterance as the relevant
message or move far away from it by following a complex inferential chain. In the
former case, the interpretation is literal; in the latter, it is non-literal. Within the best
of possible worlds, literal interpretation matches literal expression and vice-versa. In
our imperfect, human world, harmony is not pre-established; its achievement is the
moral task of the actors of communication. In a world where messages were integrally
encoded in expressions, the homo loquens would be no more than a robot devoid of
ethical dignity, the best approximation to S.Paul’s cymbalum tinniens.
92 MICHELE PRANDI
expressions can end in a contradictory act. On the other hand, a con-

tradictory expression is no obstacle towards a coherent speech act on
condition that an act of inference on the part of the addressee overturns
its essential logical property.
Applying the same line of reasoning to the relationship between
contradiction and conceptual consistency, one could be led to think that
a contradictory expression ends by conveying a consistent conceptual
content in a similar way, that is, thanks to an act of interpretation that
overturns the essential logical property of the expression. Moreover,
this is the way figures of content, and in particular metaphors, are
interpreted in texts and discourses out of inconsistent expressions.
When a metaphor is actually interpreted, the essential logical property
of the expression is overturned along the path leading from meaning
to message: an inconsistent meaning is ready to become a consistent
and coherent message thanks to a contingent chain of inferences. For
instance, when Fielding writes:
At least the Ocean, that hospitable friend to the wretched, opened her
capacious arms to receive him [Tom Jones]
he depicts an inconsistent and incoherent state of affairs. The state
of affairs is inconsistent because it attributes some human features –
arms, friendly attitude and purposeful action – to a non-human being:
to the Ocean. And it is incoherent because its content does not imme-
diately fit the narrative plot. Once interpreted by a cooperative reader,
however, this meaning is ready to be developed into a consistent and
coherent message, which is framed in words by the author himself:
Tom Jones determined to go to sea.
The idea defended here is that contradiction does not work in this
way. Of course, the textual value of a given oxymoron is the aim of
complex interpretative strategies, just as the value of any other figure,
and more generally of any actual utterance belonging to a coher-
ent text. The point, however, is that the accessibility of a consistent
conceptual content is not as such the outcome of an interpretative
process that overturns the essential logical property of the expres-
sion, but an essential logical property of the expression itself: a formal
contradiction is compatible with a consistent conceptual content. The
co-occurrence of formal contradiction and conceptual consistency is

an intrinsic property of the semantic structure of at least some kinds
of contradiction.
Let us go back to Catullus’ utterance Odi et amo. When interpret-
ing such an utterance, the reader has to make some inference in order
to imagine in what sense the poet can both hate and love the same
woman. His conclusion will probably be that the poet lives in a state of
conflict, torn between two opposite passions. In order to construct an
adequate interpretation field (Prandi 2004, Ch. 1, § 3), a professional
reader can even go farther. For instance, he can explore the philosophi-
cal background of Catullus’ attitude towards love, as Bishop10 does
when remarking that “Roman Stoicism recognizes odium [hatred] and
amor [love] as parts of the vice libido [lust]”. But this is beside the
point. One thing is to make inferences in order to identify the exact
content of an emotive state of a person in a given moment, and another
thing is the consistency of the move itself – namely, of attributing a
more or less complex emotive state to a person. Even in the absence of
contradiction, one can be led to draw some inference in order to assess
in each particular case – on the field, so to speak – in what sense and
within which limits a concept is actually applied to a given subject: for
instance, in what sense and to what extent a person can be said to love,
or to hate. In both cases, making inferences in order to find out the
contingent conditions of application of a concept to a particular subject
takes for granted – presupposes –that this concept can be applied to
the subject in a consistent way. Now, the presence of a contradictory
expression does not affect this presupposition, for both the poles of the
contradiction – hate and love - are consistent with the subject. If for
a human being it is consistent to hate and it is consistent to love, why
should it not be consistent both to hate and to love? However difficult,
the task of interpreting an oxymoron within a given text relies on a
firm, consistent conceptual ground. In the following lines, I shall pro-
vide arguments for the idea that contradiction is perfectly compatible
with conceptual consistency.
10
Bishop 1971, p. 640, quoted by Gigliucci 1990, p. 19.
94 MICHELE PRANDI
Conceptual consistency has to do with identity. However, the identity

of a being with itself is defined at two levels, that is, at conceptual and
empirical level. If such an inanimate being as the moon is seen as an
animate being, its conceptual identity is affected. If it is seen as a star, its
empirical identity is affected. The relevant level for consistency is con-
ceptual identity, which is presupposed by empirical identity. Whereas
empirical identity, which is a matter of truth, is governed by the actual
state of things, conceptual identity is governed by a system of consist-
ency criteria independent of actual experience and belonging to a natural
ontology (§ 1.4). To see the moon as a human being is inconsistent; to see
it as a star is empirically false but consistent11. On this premise, a contra-
diction is compatible with conceptual consistency insofar as it does not
challenge the conceptual identity of the discourse topic with itself.
A tautology – The moon is the moon – simply states the identity
of the subject with itself. Unlike a tautology, a synthetic predication
crosses the borders of the subject and its identity with itself in order
to predicate something about it. A synthetic predication is consistent
if the predicate confirms the identity of the subject. So is, for instance,
The moon shines: the moon that shines is still the moon; its identity
as an inanimate heavenly body is not called into question. A synthetic
predication is inconsistent if the predicate challenges the identity of the
subject. So is, for instance, The moon smiles (Blake): the moon that
smiles is no longer the moon; its identity as an inanimate being is called
into question. According to this criterion, a contradiction like The moon
shines and does not shine is a kind of consistent synthetic predication.
The moon that shines and does not shine remains the moon. It is dif-
ficult to say how it looks exactly on empirical grounds, but its concep-
tual identity as an inanimate being is not called into question. The two
opposite predicates – to shine and not to shine –are in conflict with each
other but neither is in conflict with the moon. The moon can either shine
11
Besides conflicts grounded on inconsistency, which challenge the conceptual
identity of beings, there are conflicts internal to consistent concepts, or shallow con-
flicts, which challenge the empirical identity, and are also ready to be interpreted as
metonymies, synecdoques or metaphors. To call butcher a surgeon, for instance, rises
such a kind of shallow conflict ready to be interpreted as a metaphor.
or not shine, display many degrees of intensity in its shining, and pass
from one state into the opposite during time, without stopping being the
moon. The competition between shining and its negation for determin-
ing the moon leaves the moon itself absolutely safe.
Insofar as it confirms the identity of the subject, the consistent
predication encapsulates a tautology in its foundations: “The tautology is
asserted, the contradiction is denied, by every proposition” (Wittgenstein
1960, 25.5.15). This underlying tautology can be brought to the surface
through a cumbersome but revealing reformulation: The moon shines,
for instance, can be reformulated into The moon, which is an inanimate
being, shines, a state that applies to an inanimate being. Insofar as it
challenges the identity of the subject, the inconsistent predication encap-
sulates a contradiction, which can be detected in the same way: The
moon smiles, for instance, can be reformulated into The moon, which is
an inanimate being, smiles, which presupposes that it is a human, and
therefore animate being.
If it is submitted to the same test, a contradiction of the form x is P and
non-P behaves like a consistent predication: like a consistent predication
and unlike an inconsistent one, it is built upon an encapsulated tautology.
The sentence The moon shines and does not shine, for instance, can be
developed into the form The moon, which is an inanimate being, shines,
a state that applies to an inanimate being, and does not shine, a state that
applies to an inanimate being. Like a consistent predication, a contradic-
tion of the form x is P and non-P does not affect the conceptual identity
of the subject, which is confirmed twice.
There is a point in our line of reasoning that looks somehow para-
doxical: inconsistency has at its foundations a structure – a contradic-
tion – that is compatible with a consistent content. However, the paradox
dissolves as soon as one thinks that what is correlated with inconsistency
is not the presence of an underlying contradiction as such, that is, as a
formal structure, but the substantive conceptual content it receives in
this position.
The form of contradiction that lies at the foundations of an inconsistent
predication directly opposes the subject and a predicate that negates it: x is
not x. Unlike a contradiction of the form x is P and non-P, this form of con-
tradiction is not a kind of synthetic predication, but simply the negation of
96 MICHELE PRANDI
the tautology x is x. Like a tautology – The moon is the moon – and unlike
a synthetic kind of contradiction – The moon shines and does not shine –
this form does not cross the borders of the subject. Whereas a tautology
states the identity of the subject with itself, the corresponding contradic-
tion negates it: The moon is not the moon. The formal skeleton being equal,
however, what is relevant to consistency is the conceptual content of the
underlying opposition. When it immediately involves the categorisation of
a being as such – The moon is not the moon – contradiction challenges the
empirical identity of this being. As calling into question empirical identity
presupposes conceptual consistency, this kind of contradiction is ready to
convey a consistent content (§ 4.1). When a contradiction of the same form
is presupposed at the foundations of an inconsistent predication like The
moon smiles, it challenges conceptual identity. The reason of this cleavage
is clearly not formal, because the form is the same, but conceptual: the
underlying opposition no longer involves a couple of empirical categories
but a couple of ontological ones: An inanimate, non-human being is an
animate, human being. As a purely formal structure, contradiction is inde-
pendent of consistency and does not affect it.
The above remarks allow us to conclude that conceptual consistency
is compatible with formal contradiction. If contradiction does not chal-
lenge conceptual consistency, in turn, it is because the two properties of
conceptual contents rest on different orders of lawfulness. Contradiction
breaks a formal order of lawfulness, whereas consistency is governed by
a substantive order of lawfulness. The next step will lead us to explore the
formal roots of contradiction (§ 1.3) and the substantive conceptual roots
of inconsistency (§ 1.4).
1.3. The formal ground of contradiction: syntactic structures and

formal lexical structures.
La divina scienza, che piena è di pace

[…] non soffera lite alcuna
Dante Alighieri
The syntactically based contradiction, connecting a term and its

negation – Peter is good and non-good – is a formal logical structure
in the strong sense. Besides behaving as a logical structure, it can be

detected on purely formal syntactic grounds, for the syntactic form
makes its logical structure immediately visible, irrespective of its
conceptual contents. If we extract from the last example a skeleton of
empty symbols –x is P and non-P – the logical structure of contradic-
tion remains perfectly transparent.
The lexical contradiction, connecting two opposite lexical values –
This stone is warm and cold – is a formal structure in a weaker sense,
which is kept hidden by the syntactic form. If we extract from such a
contradictory sentence as This stone is warm and cold a skeleton of
empty symbols – x is P and Q – the contradiction disappears with the
content of the symbols. The opposition and the consequent contradic-
tion have to be carved, so to speak, out of the content of the terms
involved. However, even when it is embodied in conceptual substance,
a lexical opposition remains a formal kind of structure. In order to
unmask the contradiction, one does not have to know the positive
content of the opposite terms; one has only to know that they form an
opposition. While syntactic contradiction rests on the formal configu-
ration of syntactic structures, lexical contradiction rests on the formal
organisation of lexical structures (§ 3.2.3).
1.4. The substantive ground of inconsistency: natural ontology.
Couples are wholes and not wholes,

what agrees, disagrees, the concordant is
discordant
Heracleitus
Opposites are known by one and the same

kind of knowledge
Plotinus12
The restrictions that govern conceptual consistency, that is, con-

sistency criteria, belong neither to syntactic forms nor to formal
12
Armstrong 1966, Plotinus, Enneads, I. 8. 1.
98 MICHELE PRANDI
lexical structures, but depend on the substantive conceptual content

of a shared natural ontology. It is true that any consistent predication
encapsulates a tautology and any inconsistent predication encapsulates
a contradiction. These logical structures, however, cannot be brought
to the surface through a purely formal analysis, be it syntactic or
lexical13. As they are embodied in the conceptual substance itself, the
tautologies and contradictions that lie at the foundations of consistent
and inconsistent predications can only be detected through an ana-
lytical description of the shared natural ontology, which amounts to an
analytical exposition of the system of compatibilities and incompat-
ibilities we rely upon in our consistent speech, thought and practical
behaviour.
The most immediate way of shedding light on consistency criteria
is to look at them from the epistemological observatory provided by
conflictual complex meanings (Prandi 2004, Ch. 4), that is, by mean-
ings that challenge the conceptual identity of beings. The conceptual
identity of a being can be challenged in two ways: directly, by incon-
sistent classification, and indirectly, by inconsistent connection. This
implies that consistency criteria form an interconnected system of
classificatory and relational categories (see § 3.1).
The conflict stems from classification when a being or a relation
belonging to a given category is subsumed under an inconsistent cat-
egory by a linguistic expression: a spinner is a living stone (Lawrence);
wine is vine’s dew, while poetry is mind’s sweet fruit (Pindar); the
wind is the breath of heaven (Marlowe); a smile is the real sunshine of
feeling (Charlotte Brontë).
The conflict has a relational source when beings classified in a con-
sistent way are involved into inconsistent properties or processes:
And Winter pours its grief in snow
When Autumn’s leaves are lying
Emily Brontë
Husserl (1901-1970, 4th Inquiry) distinguishes a formal, analytical sort of

13
counter-sense, which violates the laws of logic, that is, contradiction, from a material,
synthetic sort, involving the conceptual purport, that is, inconsistency.
Shadows – hold their breath

E. Dickinson
The two ways of attaining inconsistency show en filigrane the

two basic, perpendicular dimensions of natural ontology: a classifica-
tory dimension, which keeps apart the ontologically relevant kinds of
being, and a relational dimension, which imposes restrictions on the
applicability of properties and processes to the relevant kinds of being.
In formal ontological terms, the paradigmatic dimension of natural
ontology presupposes the availability of punctual classificatory con-
cepts, while the syntagmatic component requires the accessibility of
relational concepts (§ 3.2.1).
Within the classificatory component, all conceivable beings are
distributed in such conceptual categories as concrete and abstract,
animate and inanimate, human and non-human beings. This classifica-
tion, however, is not self-containing. It is not grounded on immanent
criteria, as in the case of a true taxonomy, but on criteria that are in
turn relational, because it is functional to the construction of consistent
relational patterns – namely, to the identification of processes and qual-
ities that are either consistent or inconsistent with each kind of being.
Whereas the empirical identity of a being is one and the same, its con-
ceptual identity is multi-faceted because it is relational. In our direct
experience, for instance, a tree is a tree. Looked at from the standpoint
of natural ontology, its classification is variable according to the differ-
ent classes of relations that can be applied to it. If we think of colour,
it is a physical body provided with an extended surface, like a wall, a
piece of paper, the body of a beast or of a person. If we think of percep-
tion, it is an inanimate being, like a stone or a machine. When we cut
one of its branches, for instance, we assume that it does not feel pain.
If we think of free and responsible action, it is a non-human being, like
the moon or an insect. If we are hit by one of its falling fruits, we do
not see in it an intentional act. Natural ontology is naturally anthropo-
centric. Persons enjoy a positive conceptual identity. Though owning a
physical body and sharing many functions with other living creatures,
human beings are not seen as non-animals, or non-angels, or non-gods,
but plainly as human beings. Beneath the threshold of humanity, iden-
100 MICHELE PRANDI
tity becomes less and less positive and more and more negative. A dog,
for instance, is both animate and non-human, whereas a stone is simply
inanimate. Such negative categories as inanimate or non-human cannot
be justified in view of consistent classification; they can only be justi-
fied in view of consistent relations, that is, thinking of consistent and
inconsistent processes and properties.
The relational component of natural ontology is formed by a
system of consistency criteria, called selection restrictions by lin-
guists. Selection restrictions are traditionally associated to linguistic
structures, either to syntax (Chomsky 1965, following Carnap 1932)
or to lexical structures or contents (McCawley 1970; Lakoff 1971;
Wierzbicka 1980, p. 87; Dik 1989, p. 91; Geeraerts 1991). When they
are not directly located in language, selection restrictions are seen as
cognitive models (Fillmore 1977, p. 130) or as “beliefs about the world”
(Haiman 1980, p. 345). In fact, the combinatory restrictions that gov-
ern conceptual consistency form a true grammar of concepts, which is
logically prior not only to syntactic forms and language-specific map-
pings of concepts into lexical paradigms but also to positive knowl-
edge, cognitive modelling and believing.
The formal syntactic structures of a language and the syntax of con-
sistent concepts form independent orders of lawfulness, as is shown by
the formal possibility of inconsistent complex meanings (Husserl 1901;
Prandi 1987; 2004). An inconsistent sentence – for instance The moon
smiles – breaks no formal distributional restriction. On the contrary,
it is precisely thanks to its formal scaffolding, which is insensitive to
the pressure of constructed concepts, that a sentence has the strength
to put together atomic concepts in a creative way. As it occupies the
position of grammatical subject, for instance, the moon cannot escape
the role of experiencer of dreaming. Conceptual conflict is not up to
dismantling a formal syntactic scaffolding.
The language-specific organisation of concepts in lexical para-
digms, for its part, is by definition internal to conceptual areas that
are previously assumed as consistent (Prandi 2004, Ch. 7). The lexi-
cal paradigm organising the conceptual area of killing in English,
for instance, contains such values as murder, assassinate, slaughter,
exterminate, execute, slay, butcher, and massacre. Each of these
different lexemes is submitted to language-specific lexical solidari-

ties (Porzig 1934), which impose rather arbitrary restrictions on the
kinds of being that can occur in the position of patient. Murder is
restricted to persons; “assassinate adds the restriction that the object
must be a person in a position of political importance and that the
agent has a political motive for killing. Slaughter and butcher seem
to be terms used primarily for the killing of animals for food […]
Slay is applied to humans or higher animals, overlapping somewhat
with slaughter, but it has an archaic, especially biblical, connotation.
Exterminate is usually used for intentionally killing in order to get
rid of fairly low forms of animal life, e. g. insects, or animals that
are considered pests, e. g. rats […] Massacre adds the feature that the
object consists of a group of people […] Execute is like kill, and adds
the qualification that the act is a punishment for a crime and is car-
ried out according to the laws or mores of a social group14” (Lehrer
1974, pp. 123-124). As the examples show, a language is sovereign
when imposing specific restrictions on the use of words, but it is so
on one condition: all these restrictions are internal to the borders
of consistency, and presuppose it. English, for instance, can freely
legislate about what kinds of being can be murdered, slaughtered or
massacred, but on the preliminary condition that all these beings are
mortal, and therefore living beings. It is not the task of English to
state what kinds of being can die, and therefore be killed. Language-
specific lexical structures organise in an arbitrary way consistent
conceptual areas whose external borders are drawn by consistency
criteria on independent grounds15.
14
As Lehrer points out, these restrictions are highlighted by metaphorical uses:
slaughter and butcher, for instance “are used for killing human beings in a violent
manner where the victims are treated as animals”. In a similar way, exterminate
“can be used with human victims – the Nazis’ systematic killing of Jews has been
described as extermination – and the description is effective because it emphasises
the dehumanisation of the victims”. On metaphorical uses of verbs, see Prandi 2004:
Ch. XI, § 4.2.
15
This is the criterion for distinguishing lexical solidarities and consistency
criteria, traditionally called selection restrictions. Lexical solidarities impose further
combinatory restrictions on consistent conceptual relations – only a human being
102 MICHELE PRANDI
What has been said about language holds for knowledge and cogni-
tive categorisation. The contents of both positive knowledge and cog-
nitive modelling are by definition consistent. There is no room, within
real experience, for such inconsistent beings as pregnant rocks or such
inconsistent processes as the smiling of the moon. As they highlight
typical patterns included in actual experience, cognitive models are
in turn necessarily consistent. This means that both actual experi-
ence and cognitive models are located within conceptual territories
whose external borders are previously drawn by consistency criteria.
Inconsistent beings and processes are conceivable only as complex
meanings of significant expressions, that is, as semantic structures of
the symbolic order.
Consistency criteria belong to a ground of concepts that lies deeper
than linguistic and cognitive structures because they are relied upon
as presuppositions of our practical everyday behaviour, which includes
both language and thought. Consistency criteria are part of a shared
and solid conceptual equipment that provides the conceptual constitu-
tion, so to speak, of our form of life. Before governing the consistency
of lexical structures, complex meanings and cognitive contents and
models, the shared system of consistency criteria governs the consist-
ency of our everyday behaviour. The reasons that push one to think
that The moon smiles is an inconsistent complex meaning that meets
neither experience nor conceptual modelling are the same that prevent
one from addressing statements, questions and orders to the moon.
Consistency criteria certainly form, as Wittgenstein (1969, p. 97) puts
it, “the river-bed of thoughts”, but only insofar as they form the river-
bed of the whole game of life, tacitly governing the consistent behav-
iour of human beings (Prandi 2004, Ch. 8).
The categorisation of beings that is assumed by natural ontology is
of no theoretical or cognitive import; it is not framed in propositions
but tacitly presupposed by consistent action. A man using a stone to
of political importance, for instance, can be assassinated – whereas consistency

criteria circumscribe from outside the area of consistency: only living beings can
be killed.
drive in a nail practically categorises the stone as an inanimate instru-

ment. He does not explicitly define the stone as an instrument, but
simply uses it as if it were one. An individual who argues with another
practically categorises the addressee “as a person, a rational agent”
(Dennett 1969, p. 177), who can evaluate different options and make
rational choices. If trees are classified as non-human beings within our
shared ontology, it is insofar as people do not behave towards them as
they behave towards human beings16.
As they are located beneath the threshold of both language-specific
lexical structures, positive knowledge and cognitive modelling, con-
sistency criteria are not sensitive to the empirical properties of beings,
but only to their ontological possibility. The empirical predicates
actually or typically displayed by the different kinds of beings – for
instance, hate and love – provide lexical paradigms, and more specifi-
cally oppositions, with their conceptual purport. What is relevant to
consistency, however, is not either of the opposite values of a paradigm
– for instance hate or love – but the whole conceptual space spanning
between them. As Sommers (1963, p. 160) sums up, “The ontologist is
interested in categories; he is, qua ontologist, not interested in whether
a thing is red or whether it is green but in whether it is coloured. Even
this is not altogether accurate: he is interested in its character of being
coloured or colourless”, that is, in its aptitude to display a colour.
The immediate consequence of ontological categorisation is that
the distinction between true and false predications, which is relevant
16
Searle (1983, pp. 158-159) defines Background the network of concepts that
underlies consistent behaviour. Realism is the most significant example. Before being
an explicit philisophical claim, realism is a practical attitude, which is “part of the
Background”: “My commitment to ‘realism’ is exhibited by the fact that I live the way
I do, I drive my car, drink my beer, write my articles, give my lectures and ski my
mountains […] My commitment to the existence of real world is manifested whenever
I do pretty anything […] This is not to say that realism is a true hypothesis, rather it
is to say that it is not an hypothesis at all, but the precondition of having hypotheses”.
The practical relevance of a presupposed layer of complex conceptual structures
shows that shared concepts do not reduce to either cognitive contents or cognitive
structures. Beyond their different categorial contents, it is a different attitude on the
part of the subjects that draws a sharp line between presupposition and cognition.
104 MICHELE PRANDI
not only to cognitive contents but also to cognitive modelling, is of

absolutely no importance, because it is located inside the borderline
that consistency criteria draw from outside: “whenever a predicate P
is significantly applicable to a thing, then is its complement non-P […]
Thus, any predicate P can be constructed as |P| or ‘the absolute value
of P’, by which we mean that P spans the things which are either P
or non-P but does not span things which are neither P nor non-P. For
example, if P = philosopher, then |P| defines the class of things that are
either philosophers or non-philosophers”, that is, human beings. “In the
class of things that are |P| are Bertrand Russell and Cleopatra, but not
the Empire State Building17” (p. 159).
If a contradiction is allowed to convey a consistent conceptual
content, it is for precisely this reason: given a kind of being, both its
17
The distinction between lexical opposition and negation has some ontological
implications. A lexical opposition defines a homogeneous conceptual space. Death,
for instance, is as positive a reality as life itself, and has the same range of consistent
application, that is, animate beings. Thanks to this, the correlative concepts life and
death circumscribe the whole consistent conceptual area of life, as distinct from the
realm of non-living beings. The correlation between a term and its negation, on the
contrary, is asymmetric: the two terms have different ranges of consistent application.
While living defines a positive and homogeneous concept, the content of non-living
defines a residual and non-homogeneous area, which includes the deprivation of life
– internal negation – as well as the pure lack of life – external negation. A stone and
a corpse, for instance, can both be defined as non-living entities. The two statements,
however, do not have the same content. When a corpse is defined as non-living, it is
understood that it has been the body of a living person. When a stone is defined as
a non-living being, what is meant is that the stone is outside the consistent predica-
tion range of life, and hence of death. Whereas the corpse is a body deprived of life,
the stone simply lacks it. The distinction between the pure lack of a property and its
deprivation by an antagonistic force (Kant 1763, p. 217) was first drawn by Aristotle
(The Categories, 12a). Within Sommer’s argument, the denotation range of the
negative predicate has to be taken as being internal to the relevant conceptual area:
non-philosopher, for instance, may mean scientist or carpenter but not car or star;
non-living means dead creature, and not inanimate being. It is worth noting that in
real texts this distinction is practically neutralised, because the presence of consist-
ent arguments restricts the relevant conceptual area. Catullus’ I hate and love and
Anacreon’s I love and I do not love, for instance, describe two conflicts that, however
different in content (see § 4.3), are both located within a human soul.
consistent and inconsistent predication range span between couples

of opposite terms. A human being, for instance, is consistent with the
whole conceptual space spanning from love to hate, whereas a heav-
enly body is not. If it is consistent for Catullus to love Lesbia, it is
consistent for him to hate her too. If it is inconsistent for the moon to
love Lesbia, it is inconsistent for it to hate her too. For the same reason,
contradiction being equal, Catullus hates and loves Lesbia depicts a
consistent state of affairs, whereas The moon hates and loves Lesbia
depicts an inconsistent one. Of course, it is not so easy to define exactly
what kind of emotive state is actually attributed to Catullus. Whatever
it is, however, this state is necessarily located somewhere within the
consistent conceptual space spanning from love to hate. Accordingly,
the difficulty is purely empirical, not ontological; it is relevant for
truth, not for consistency. A contradiction that makes two opposite
predicates clash, both of them consistent with the subject, does not
challenge consistency.
1.5. The consistent contents of contradiction I: complexity, conflict

and change.
Vos cogitastis de me malum et Deus vertit

illud in bonum
Genesis
War is the father of all and the king of all

Heracleitus
Vinceris aut vincis, haec in amore rota est

Propertius
If it is not relevant for consistency whether x is P or non-P – wheth-

er Catullus loves or hates Lesbia – it is because the whole conceptual
space spanning from P to non-P – from love to hate – is consistent with
the subject. If so, the conceptual consistency of contradiction is only a
particular case of a more general fact – namely, the fact that it is con-
sistent to engage two opposite concepts in the determination of a given
subject. There are three orders of good reasons that justify this fact.
106 MICHELE PRANDI
First, things are complex, and within the structure of complex

phenomena a concept rarely goes without its opposite. Beauty is never
without shade, nor happiness without some bitter drops of sorrow: All
evils are to be considered with the good that is in them (Defoe).
Complexity easily turns into conflict. This happens, in particular,
when a human soul is torn between two opposite attitudes: Odi et amo;
Amor mi sprona in un tempo et affrena, / assecura et spaventa, arde et
agghiaccia, / gradisce et sdegna, a sé mi chiama et scaccia (Petrarch:
Love spurs and holds me back at the same time, / and frightens and
reassures, freezes and burns, / is kind and rude; he calls me, throws me
out18); I both wished and feared to see Mr Rochester (Charlotte Brontë);
Vorrei e non vorrei19 (Da Ponte: I should like to, yet I shouldn’t).
Things, and above all human condition and feelings, are bound to
change, passing from one conceptual pole to the opposite20. Owing to
this, if the presence of a temporal gap can be inferred, an utterance
sharing the outer structure of a contradiction can easily be interpreted
as if it were a consistent description of a change in time: Io chiudo ed
18
Musa 1996.
19
Catullus’ oxymoron comes back in Renaissance poetry: D’amore et d’odio in
qual guisa si mova / il vario affetto in me, no’ ‘l saprei dire (Cariteo: In what guise
do love and hate my changing spirit move, I cannot say). More generally, oxymoron
is the very mark of love whithin the mainstream of Western poetry (Gigliucci 1990;
2004). Gigliucci (2004, pp. 12-13) quotes Romeo’s outburstings – O heavy lightness!
serious vanity! / Mis-shapen chaos of well-seeming forms! / Feather of lead, bright
smoke, cold fire, / sick health! / Still-waking sleep, that is not what it is!/ This love
feel I, that feel no love in this – and Longinus’ commentary on Sappho’s fr. 12: “She
feels contradictory sensations, freezes, burns, raves, reasons, so that she displays not
a single emotion, but a whole congeries of emotions. Lovers show all such symptoms,
but what gives supreme merit to her art is, as I said, the skill with which she takes up
the most striking and combines them into a single whole” (Innes 1995).
20
J. Burckhardt suggests with a telling allegory that the conflicts and tensions
that affect the course of history dissolve in harmony when looked at from the histo-
rian’s distant and high standpoint: Von einem hohen und fernen Standpunkt aus, die
der des Historikers sein soll, Klingen Glocken zusammen schoen, ob sie in der Naehe
disharmonieres oder nicht: Discordia concors (Listen from high and afar, from a
place which should be that of the historian, bells make a wonderful sound even if,
from close by, they may not always be in harmony: Burckhardt, p. 142).
apro, celo e manifesto (Bracchi: I [that is, a door] open and close, con-
ceal and reveal). Complexity, conflict and change affect the empiri-
cal content of the predicate, but they do not challenge the conceptual
identity of the subject.
Last but not least, formal contradiction is used, from Heracleitus
onward, as the only consistent way for speaking about God. If God
contains all the conceivable determinations, he necessarily spans the
opposite ones: God is day and night, winter and summer, war and
peace, surfeit and hunger21. If God is coincidentia oppositorum, theol-
ogy can only be conceived of as docta ignorantia (Cusanus). The first
oxymoron theatrically suggests that God escapes any finite conceptual
determination (Ossola 1977); the second, that it is impossible to speak
of God without falling into contradiction.
If complexity, conflict and change are constitutive of the finite
existence of things and creatures, the consequence is that the consist-
ent description of empirical properties and processes involving things
and creatures not only is not incompatible with, but actually requires
the co-occurrence of two opposite predicates. This, however, does not
imply that contradiction is the logical form of complexity, conflict and
change. Contradiction is simply an option, which lends to the expres-
sion the appeal of logical extremism.
Though involving opposite concepts, complexity, conflict and
change form consistent states of affairs that can be framed in non-
contradictory expressions. As Kant (1763, p, 211) points out, a contra-
diction is a formal property of some linguistic expressions, whereas
conflict involves the conceptual substance of things22. Complex things
21
Jones 1959, Fragment XXXVI.
22
Kant (1763, p. 211) draws a sharp distinction between contradiction, a prop-
erty of expressions, and complexity and conflict, to be found in reality. According
to him, a logical contradiction “consists in the fact that something is simultaneously
affirmed and denied of the very same thing. The consequence of the logical conjunc-
tion is nothing at all (nihil negativum irrepraesentabile), as the law of contradiction
asserts. A body which is in motion is something; a body which is not in motion is
also something (cogitabile); but a body which is both in motion and also, in the very
same sense, not in motion, is nothing at all”. A conflict in reality – a ‘real opposi-
tion’ in Kant’s terminology – “is that where two predicates of a thing are opposed
108 MICHELE PRANDI
can be described through non-contradictory expressions, and contra-

diction is no more than an accident in the expression of complexity,
conflict and change. If this is true, what makes a contradictory utter-
ance an attractive form of expression is not so much the contradiction
itself, but the simultaneous presence within a simple expression of the
opposite terms that circumscribe the conceptual space of complexity,
conflict and change. Now, contradiction is both the formally simplest
and conceptually extreme form of co-occurrence of two opposite con-
cepts, but it is not the only form.
In order to prevent the simultaneous use of two opposite terms from
falling into contradiction, it is enough to articulate their relation with
appropriate linguistic devices. The following utterance, for instance,
frames a complex state of affairs in consistent words. If we take away
the two parallel explanations, the result is an oxymoron, That sentence
was true and false:
That sentence was true – because I had really said that I could not speak about
that book – and false, because it indirectly but vigorously put about a sup-
posed opinion of mine which actually did not exist (Magris).
A typical way of avoiding contradiction while expressing a complex
or conflictual state of affairs is to coordinate two opposite concepts as
to each other, but not through the law of contradiction. Here, too, one thing cancels
that which is posited by the other; but the consequence is something (cogitabile). The
motive force of a body in one direction and an equal tendency of the same body in the
opposite direction do not contradict each other; as predicates, they are simultaneously
possible in one body. The consequence of such an opposition is rest, which is some-
thing (repraesentabile)”. For Hegel, who identifies real and rational, contradiction is
the logical form of complexity, conflict and change. Accordingly, as Poublanc (1991,
note 15) points out, he uses the term contradiction to denote complexity and conflict,
that is, Kant’s ‘real opposition’. Moreover, Hegel’s use of the term contradiction is
so generous as to include in its denotation range both syntagmatic combinations
and paradigmatic oppositions: “contradiction is […] immediately represented in the
determinations of relationship. The most trivial examples of above and below, right
and left, father and son, and so on ad infinitum, all contain opposition in each term.
That is above, which is not below; ‘above’ is specifically just this, not to be ‘below’,
and only is, in so far as there is a ‘below’; and conversely, each determination implies
its opposite” (Hegel 1812-1813, p. 441).
arguments of a predicate that plainly describes the way they co-occur

in things.
As they can consistently merge and conflict, opposite feelings can
consistently be said to merge and conflict. When Odisseus’ nurse
Euriclea suddenly recognised her master, Then upon her heart came
joy and grief at the same moment23; The pleasures of heaven are with
me and the pains of hell are with me (Whitman); Joy and sadness /
are merged / in this spring air (Bacher); Zwei Seelen wohnen, ach,
in meiner Brust (Goethe: Two souls live in my breast); Così sol d’una
chiara fonte viva / move ‘l dolce et l’amaro ond’io mi pasco (Petrarch:
So from one clear and living font alone / there springs the sweet
and bitter that I feed on). In the following examples, contradiction
is avoided through a metaphorical predicate that makes direct refer-
ence to a conflict: Amore e odio in mezzo hanno il cor mio (Cei: Love
and hate battle over my heart); Such civil war is in my love and hate
(Shakespeare).
Opposite properties of one and the same being take turns in time;
therefore, the plain description of change connects opposite concepts
without contradiction: Cold things become warm, warmth cools, mois-
ture dries, the parched gets wet (Heracleitus24); Mortal men are unfamed
and famed alike, and named and unnamed, by the will of great Zeus.
For easily he strengthens, and easily he crushes the strong, easily he
diminishes the conspicuous and increases the inconspicuous, and eas-
ily he straightens the crooked and withers the mainly (Hesiod25); If she
runs away, soon she shall pursue; if she does not accept gifts, why, she
shall give them instead; and if she does not love, soon she shall love even
against her will (Sappho26); Amor […] or mi tene in speranza et or in
pena, / or alto or basso il meo cor lasso mena (Petrarch: Love […] with
hope he holds me now and then with grief, / now high now low he leads
my weary heart27); Gelo, e poi sento / l’alma avvampar / e in un momento
23
Murray 1995, Od. 19.
24
Jones 1959, Fragment XXXIX.
25
Most 2006.
26
Campbell 1982.
27
Musa 1996.
110 MICHELE PRANDI
/ torno a gelar (Da Ponte: I freeze, and then I feel my soul aflame, and in
a moment, I turn cold again); To day we love what to morrow we hate; to
day we seek what to morrow we shun; to day we desire what to morrow
we fear […] Such is the uneven state of human life (Defoe)
As a linguistic picture of complexity and conflict, oxymoron can be
considered a logically radical alternative to analytical description – to
a description capable of distributing the characterisation of a subject
among two opposite terms in a non-contradictory way. Many instances
of analytical description can be found in narrative texts:
Tess had never before known a time in which the thread of her life was so dis-
tinctly twisted of two strands, positive pleasure and positive pain (Hardy).
[…] and the respect he felt for her high rank, and his veneration for her as
his patroness, mingling with a very good opinion of himself, of his authority
as a clergyman, and his rights as a rector, made him altogether a mixture of
pride and obsequiousness, self-importance and humility (Austen).
He was standing two yards from her with his mind full of contradic-
tory desires and resolves – desiring some unmistakable proof that she loved
him, and yet dreading the position into which such a proof might bring him
(George Eliot).
With its consistent elaboration of the opposite predicates, the last
passage avoids both contradiction – I both wished and feared to see Mr
Rochester (Charlotte Brontë) – and tautology: Catherine sometimes […
] hoped or feared that she had gone too far (Austen).
The two ways of expressing complexity and conflict – analytical
description and oxymoron – typically coexist in texts. Sometimes, a
description holds as an explanation of a previous oxymoron:
I wished, yet feared, to find him. I felt the terrible news must be told, and I
longed to get it over; but how to do it, I did not know (Emily Brontë).
I both wished and feared to see Mr. Rochester on the day which followed
this sleepless night. I wanted to hear his voice again, yet feared to meet his
eye (Charlotte Brontë).
Sometimes an oxymoron holds as a sort of recapitulation of a pre-
vious description. In the following example, a consistent description
which distributes the opposite predicates among the different sides

of a split soul turns into contradiction when the subject recovers her
unity:
He had uttered precisely what her soul yearned for, but which her reason
dreaded […] and she was both frightened and made happy thereby (Tolstoy).
Contradiction has on its consistent counterpart the advantage of dra-
matically doubling the conflict in things with a conflict in words. Its
appeal is so strong that it can even happen that a consistent expression
is framed in such a way as to mime the outer appearance of oxymoron:
Proinde saepe ac multum cogitavi si quis inveniri posset incantus qui
me absentem tibi praesentem et longinquum propinquum et tacentem
loquentem faceret (Guarino to Lionello D’Este: Thus, I often wondered
a lot if some spell could be found that makes me be for you present
while absent, nearby while faraway, silent while speaking).
If we take all these facts into account, the question about the “rela-
tionship between contradiction principle and correlation principle”
(Bottiroli 2006, p. 161) is reduced back to its real scale. First, formal
contradiction does not imply conceptual inconsistency. Moreover, con-
tradiction is not the logical form of complexity, conflict and change, for
the opposite concepts required for the adequate expression of all these
processes can co-occur without contradiction, according to the identity
principle. Finally, and for the same reasons, interpreting a contradic-
tion amounts to conceiving a consistent thought. This shows that the
identity principle, or principle of non-contradiction, broken within the
expression, is always active in thought as a normative criterion. When
it does not immediately inspire the speaker’s words and actions, it is
bound to rule the addressee’s behaviour.
1.6. The consistent contents of contradiction II: subjects in conflict.

Talché morendo morte, alfin in morte
la vita si converte, e morte in vita
T. Tasso
The expression of complexity, conflict and change in things is

certainly the most revealing function of contradiction to be found
112 MICHELE PRANDI
in real discourse. However, if focus is displaced from the interac-

tion between the static systems of long-lasting ontological categories
and changing empirical data towards the interpersonal dimension
of speech, a supplementary space for consistent use of contradic-
tion opens – namely the expression of conflicts between persons or
even between Weltanschauungen. In this case, the responsibility of
subscribing the opposite concepts is not attributed to one and the
same speaker, but to two different subjects, who see the same facts
in opposite ways28.
In some cases, contradiction simply documents an occasional con-
flict between personal opinions. A person, for instance, may consider
stupid or easy what ordinary people see as clever or difficult: It always
seemed to me a sort of clever stupidity only to have one sort of talent
(George Eliot); occupations médiocres qui amusent l’intelligence par
des difficultés faciles (Flaubert: commonplace occupations, which
amuse the mind with facile difficulties). When reporting another per-
son’s speech, a speaker can directly display his mocking disagreement:
Les législateurs des pays membres se hâtent lentement d’harmoniser
leurs législations29 (Journal de Genève: Lawmakers from the Member
States slowly hurry to harmonise their legislation).
In more significant cases, the conflict involves incompatible
Weltanschauungen, typically competing philosophical pillars of moral
behaviour. In Sophocles’ tragedy, Antigones says to Ismenes: I shall
have accomplished a blessed crime. Her action, which is a crime
28
These are instances of polyphonic speaking, implying the reference to distinct
responsible subjects: cf. Ducrot 1980; Mortara Garavelli 1985; Landheer 1996, § 3.
29
The mix of antiphrasis and polyphony documented by such uses of oxymoron
is the same to be found in irony (Perrin 1996). The difference is in the division of
labour between speaker and addressee, who, in case of irony, coincides with the
quoted speaker. Whereas oxymoron directly opposes the antithetic view to the quoted
statement, irony simply echoes the latter, leaving to the addressee the task of turning
its content upside down in order to meet the speaker’s thought. If the same person
who wrote Lawmakers from the Member States slowly hurry to harmonise their
legislation had ironically addressed one of these lawmakers, he would simply have
said: So, you hurry up, letting him the task of drawing from the uttered meaning the
opposite message.
according to the tyrant’s law, is blessed in the light of common moral

sense. What looks as wisdom to the unbelieving philosopher is folly to
the eyes of faith: Insanientis dum sapientiae / consultus erro (Horace:
While I wander learned in a raving wisdom). What looks as death
to human judgement is life to the eyes of the soul: the philosopher
Campanella addresses his body as Morte viva (Living death). According
to Cicero, a friend’s viewpoint turns into good everything the layman
sees as evil: Et absentes adsunt, et egentes abundant, et imbecilles
valent, et, quod difficilius dictu est, mortui vivunt: tantus eos honos,
memoria, desiderium prosequitur amicorum (And those absent are
present, the poor are rich, the weak are strong, and the dead are alive,
which is the hardest thing to say: so much honour, memory and regret
come along with them on the part of their friends). When the Italian
bishop and baroque poet Petrucci bursts out Che dunque è questa fede!
/ È luce oscura, è cecità che vede (What is this faith! It is dark light, it
is seeing darkness) two voices are probably in conflict within his soul:
faith and human reason.
As it happens at conceptual level, if the conflict is explicitly
described, the contradiction disappears: Sapientia […] huius mundi
stultitia est apud Deum (St. Paul: This world’s wisdom is foolishness
with God); Folle all’occhio mortal del basso mondo / Saggio al Senno
divin dell’alto popolo (Campanella: Foolish in the eyes of the lowly
world / Wise to the divine judgement of the high people); If their praise
is censure, your censure may be praise (Austen).
A conflict of viewpoints internal to one and the same subject is the
split between being and seeming, which receives consistent expres-
sion if a predicate of seeming connects the opposite concepts: Galilei’s
telescope is un ammirabile stromento / per cui ciò ch’è lontan vicino
appare (Marino: an admirable instrument, through which that which
is far seems near);
El ancho campo me parece estrecho, The wide field to me seems narrow,
la noche clara para mí es escura, the clear night for me is dark,
la dulce compañía amarga y dura, the sweet company bitter and hard
y duro campo de batalla el lecho and hard battlefield my bed.
Garcilaso de la Vega
114 MICHELE PRANDI
2. A syntactic typology of contradictio.
Le dolce acerbe cure che dà Amore

A. Poliziano
There are three main ways of syntactically connecting opposite

terms into a relationship of contradiction: coordination, predicative
relation and modification.
The coordination of two opposite predicates, jointly applied to a
given subject, is the more open form of contradiction, fit for any couple
of term of equal rank: Odi et amo (Catullus); Ardo et son di ghiaccio
(Petrarch: I burn, and yet I freeze); I both wished and feared to see Mr
Rochester (Charlotte Brontë); Vorrei e non vorrei (Da Ponte); She was
both frightened and made happy thereby (Tolstoy). This form conveys
a synthetic, though contradictory predication. A variant of this form is
the repetition of the subject, which ends in telling parallelism: Io son
prigionia / io son libertà (Bracchi: I [that is, a door] am captivity / I
am freedom).
The predicative relation is an exocentric construction connecting a
subject and a predicate into a higher order structure – a sentence. In the
presence of a verbal predicate, the verb either denotes the opposite of
the subject – To give away yourself keeps yourself still (Shakespeare)
– or implies it: Frost itself as actively doth burn (Shakespeare). When
the opposite concepts are connected through a copula, the predicate
is either a noun – This stone is not a stone; This battle is not a battle;
Love is not love / Which alters when it alteration finds, / Or bends with
the remover to remove (Shakespeare) – or a non-finite verbal form – To
give away yourself is to keep yourself – or an adjective. The adjective
either is predicated of a subject that lends a nominal expression to the
opposite property – His stupidity is clever – or challenges an inherent
property of the subject noun: Jugum meum suave est, et onus meum
leve (Matthew: My yoke is sweet and my burden light). In its most
direct form, this kind of contradiction reduces itself to the negation of
a tautology, and therefore is not up to convey a synthetic predication:
This stone is not a stone, for instance, simply overturns the tautol-
ogy This stone is a stone. As we have already remarked, this kind of
contradiction lies at the ground of inconsistent predication, whereas

its tautological counterpart supports consistent predication, including
synthetic contradiction.
The relation of modification connects two opposite terms belong-
ing to different ranks – a head noun and a modifying adjective, or
a verb and a modifying adverb – into an endocentric subordinative
structure. Contradiction grafts on modification in two ways. The
opposite terms either jointly modify the same head – Le dolce acerbe
cure che dà Amore (Polizian : Love’s sweet bitter cares) – or, more
typically, are distributed between head and modifier: dulcem […] ama-
ritiem (Catullus: sweet bitterness); rerum concordia discors (Horace;
Ovid: discordant concord of things); cette obscure clarté qui tombe
des étoiles (Corneille: that obscure clarity that falls from the stars);
Todo era un silencio sonoro30; Formé su cordaje con mi vida muerta
(García Lorca: There was a resounding silence; I made its strings [of
the lyre] with my perished life); il muto grido dell’inviolata natura
(Fenoglio: the mute scream of unviolated nature); It always seemed
to me a sort of clever stupidity only to have one sort of talent (George
Eliot); Piangendo rido e sospirando godo (Bernardo Pulci: Weeping
I laugh and sighing I rejoice); Lawmakers from the Member States
slowly hurry to harmonise their legislation (Journal de Genève); ‘Twas
not my blame – who sped too slow (Dickinson). In some cases – for
instance in clever stupidity – the head and the modifier directly express
the opposite concepts. In others – for instance in to speed slow – the
modifier negates an inherent property of the head. In Muda la admi-
ración, habla callando (Góngora: Mute admiration speaks tacitly) the
verb is challenged twice: directly by a modifying gerund – callando –
and indirectly by a modifier of the subject: muda.
30
This theme occurs very often in García Lorca’s poetry: Silencio musical
(Musical silence); Oye el fluir de los grandes ríos que pasan en silencio (He [the
poet] hears the great rivers flowing silently); Crujen en silencio los huesos de las
montañas (The bones of the mountains creak in silence); El silencio tiene su música,
pero el sonido tiene la esencia de la música del silencio (Silence has its own music,
but the sound has the essence of the music of silence).
116 MICHELE PRANDI
An extreme case of the head-modifier form is the compound adjec-

tive: for instance, the Italian adjective agrodolce31 (bittersweet) or
Hölderlin’s creation traurigfroh (sadhappy):
Und der Jüngling, der Strom, fort in die Ebne zog,
Traurigfroh, wie das Herz, wenn es, sich selbst zu schön,
Liebend unterzugehen,
In die Fluten der Zeit sich wirft.
While away to the plain journeyed the youthful stream,

Sadly-glad as a heart, which, for itself too fair,
Longing swiftly to perish,
Plunges into the tides of time32.
Beneath the asymmetric surface of all these expressions, a sym-
metric conceptual opposition can be recovered.
More elaborate forms, distributing the correlative concepts among
less predictable positions or within more complex relationships can be
found in texts. In any case, for a contradiction to occur, it is essential
31
Unlike metaphor, metonymy and synecdoche, oxymoron enjoys a limited lexi-
cal creativity. First, the lexical valorisation of oxymora seems restricted to adjectives
– for instance Italian agrodolce (bittersweet), chiaroscuro, tragicomico – and nouns
of processes or qualities: for instance tragicommedia (tragicomedy). Moreover, oxy-
moron does not give birth to a new sense of an old word, contributing to polysemy,
but it simply combines two opposite values into a new compound word: for instance
chiaroscuro or agrodolce (Lepschy 1981, p. 196). Finally, as it denotes a complex
property merging the opposites, the compound word remains necessarily transparent.
A lexical phenomenon connected with opposition, but not with contradiction, is so
called enantiosemy (Lepschy 1981; 1989), which is documented both in diachrony, as
a shift of a word from a meaning to its opposite, and in synchrony, as a coexistence
of two opposite meanings of the same lexeme. An example of shift is Italian feriale
applied to a day, whose meaning has passed form “holyday” (lat. feria), to “week and
working day”. An example of coexistence is French sacré, meaning both holy and
blaspheme, as documented by the joke Rossini wrote on the manuscript of his Petite
Messe Solennelle: Bon Dieu, la voilà terminée, cette pauvre petite messe. Est-ce bien
de la musique sacrée que je viens de faire, ou bien de la sacré musique? (Dear God,
here it is finished, this poor little Mass. Is this sacred music which I have written or
music of the devil?).
32
Leishman 1954.
for an underlying correlation between two opposite concepts to be

detectable: A deep, autumnal tone, / Sweet though in sadness (Shelley);
Primitivo said with an absolute devoutness of blasphemy (Hemingway);
Nella mia infedeltà sono sempre stata fedelissima (L’Espresso: In my
infidelity I always remained loyal); Occidis nos, ne moriamur abs te
(St. Augustine: You bring death upon us, so that we will not die apart
from you); Para vivir un año es necesario / morirse muchas veces
mucho (Angel Gonzales: To live a year one has to die many times); You
cause him to die daily a living death (Hawthorne); They live unwooed
and unrespected fade / die to themselves (Shakespeare); Negli abissi di
luce erro all’oscuro (Melosio: In the abyss of light, I wander in dar-
kness); Sobre el olivar hay un cielo hundido / y una lluvia oscura de
luceros fríos (García Lorca: Over the olive grove a sunken sky / and a
dark rain of cold bright stars); If he were to start clutching at straws,
hoping against hope for a way out, then even that would be lost (British
National Corpus). In the expression el amplio elemento en que moran /
los ensueños despiertos (García Lorca: the wide element where awaken
dreams dwell), the participial adjective – despiertos, awaken – shifts
from its syntactic partner – the noun ensueños, dreams – to the dream-
ing human subjects, summing up oxymoron and hypallage, or oblique
modification (Prandi 2004, pp. 141-143).
The head-modifier structure – for instance clever stupidity – is the
most highly praised in literary texts, and therefore the most typical
form taken by oxymoron, the figure of contradiction par excellence33.
33
The different forms of contradictory utterance are seen as different kinds of
figure by some scholars. Cellier 1965, for instance, uses the term antithesis to refer
to the coordinative form – the type Odi et amo – and restricts the term oxymore to
the subordinative form: see also Molinié 1992, p. 235. Morier (1961) distinguishes
between oxymore – the direct contradiction, and in particular its subordinative form
– for instance symphonia discors – and paradoxisme, which “avoids the abrupt con-
frontation of oxymoron” keeping a certain distance between the opposite terms, as for
instance, Rétablit son honneur à force d’infamie (Boileau). The term paradoxisme
is Beauzée’s translation of the Greek term oxymoron (Diderot – d’Alembert 1772,
“Paradoxisme”). This attitude is typical of the taxonomic attitude of ancient rhetorics,
which hides the essential conceptual and grammatical properties of figures under a
proliferation of labels connected with marginal differences.
118 MICHELE PRANDI
The coordinative form – for instance, Odi et amo – possesses the

epistemological advantage of transparency: as it jointly attributes two
opposite predicates to the same subject, it both directly displays the
essential logical structure of contradiction and conveys a true synthetic
predication.
3. Kinds of concept and kinds of opposition.

3.1. Paradigms and oppositions: negation and lexical paradigms.
There is no contradiction in speech without underlying opposi-
tion. Contradiction is an instance of syntagmatic relation, which is a
relationship in praesentia between concepts co-occurring in actual
speech, defined ‘both-and’ function by Hjelmslev 1943. Opposition is
a kind of paradigmatic relation, or correlation, which is a relationship
in absentia involving virtual substitutes, defined ‘either-or’ func-
tion by Hjelmslev. The terms of a paradigm are in competition at the
moment of occupying a given position within a syntagmatic relation.
Given such a context as Mary is -, for instance, happy and sad are in
competition for qualifying the subject: they are virtual substitutes in
paradigmatic correlation.
Opposition is certainly a paradigmatic relation between virtual sub-
stitutes; however, not all paradigmatic relations are oppositions, which
implies that not all syntagmatic connections in praesentia of virtual
substitutes rise a contradiction. Our next point will be to isolate oppo-
sition among the different kinds of paradigmatic correlation, in order
to isolate contradiction among other kinds of conflict.
The first relevant distinction to be made on this behalf is that
between lexical paradigms and paradigms including a term and its
negation. The paradigms containing a term and its negation necessar-
ily form an opposition in the presence of any kind of concept: bad vs
non-bad; walk vs don’t walk; to be a rose vs not to be a rose34. This
does not hold for lexical paradigms.
34
The negation of a classificatory noun only makes sense in predicative position:
for instance, This is not a rose.
As we have remarked above (§ 1.3), lexical opposition is as formal

a structure as opposition based on negation, that is, a structure that
occurs identical in association with different conceptual contents and
that can be identified and described independently of them. The differ-
ence is that the formal structure of lexical oppositions is not immedi-
ately visible through the formal skeleton of the expression, but has to be
identified among many different kinds of lexical correlations. Besides
oppositions, lexical correlations include both substantive correlations,
which cannot be isolated from conceptual contents as independent for-
mal structures, and formal correlation that are not oppositions. Owing
to this, the quest for oppositions among lexical paradigms involves two
steps: first, to identify formal correlations on the background of sub-
stantive ones, and then, among formal correlations, oppositions.
The availability of formal lexical correlations depends in the first
place on the nature of the involved concepts.
Concepts that classify beings, that is, classificatory concepts, form
paradigms of virtual substitutes but they are not involved in formal
correlations, and a fortiori in oppositions. Such concepts as Rose
and violet, for instance, are in paradigmatic correlation. Given such
a context as This flower is a -, for instance, the two terms are virtual
substitutes, and as such cannot co-occur. In spite of this, they do not
form an opposition, in the first place because they do not belong to a
closed paradigm of interconnected values but to an open series devoid
of any independent formal structure.
Concepts that involve beings in processes or ascribe properties
to beings, that is, relational concepts, are open to formal lexical cor-
relations. Formal lexical correlation, however, is a necessary but not a
sufficient condition for opposition. Whereas all oppositions are neces-
sarily formal lexical correlations, there are many formal lexical cor-
relations that are not oppositions. Happy and sad, for instance, form an
opposition, whereas walk, run, hop, skip and crawl do not: they form a
purely differential paradigm35.
35
Other cases of formal correlations that are not oppositions are hyponymy and
synonymy. A hyponym and its hyperonym – for instance murmur and speak – are
120 MICHELE PRANDI
A supplementary source of complexity to be faced on the quest

for opposition and contradiction is polysemy. When the meaning of
a lexeme splits into a constellation of independent tough interrelated
senses, each sense enters into a distinct network of lexical correlations,
including oppositions.
In the following pages, we shall examine all these relevant points.
First, we shall make explicit the distinction between classificatory and
relational concepts (3.2.1). Then, we shall look deeper into the formal
properties of lexical paradigms. After drawing the distinction between
closed paradigms of relational concepts, which contain formal correla-
tions, and open series of classificatory concepts, which do not (3.2.2),
we shall focus on closed paradigms in order to identify oppositions
against the background of differential paradigms (3.2.3). Finally, we
shall examine the relevant formal kinds of opposition (3.2.4) and the
relationship between polysemy, opposition and contradiction (3.2.5).
3.2. Lexical paradigms and oppositions.

3.2.1 Classificatory and relational concepts.
What’s in a name? That which we call a rose

By any other word would smell as sweet
W. Shakespeare
The concepts forming our shared patrimony belong to two different

and complementary kinds. There are concepts that subsume instances
virtual substitutes in paradigmatic correlation. In spite of this, they do not form an

opposition, because a hyponym and its hyperonym have different ranks. Synonyms
– for instance dread and fear or murmur and mutter – are virtual substitutes too, but
they do not form an opposition because they are jointly located on the same side of an
opposition, or, more generally, of a lexical relation: murmur and mutter, for instance,
jointly contrast with shout. Owing to this, any synonym is ready to enter into con-
tradiction with any synonym located on the other side of the opposition. The syn-
tagm gentil descortesía (graceful discourtesy), for instance, is no less an oxymoron
than cortés descortesía (kind discourtesy) or cortés grosería (kind rudeness): Con
gentil descortesía / mueve el viento la hebra voladora (Góngora: With its graceful
discourtesy / the wind blows ’round the swiftly flying strands).
of beings under categories: they can be defined classificatory concepts.

And there are concepts that impose relations, that is, properties or
processes, on instances of beings, or involve instances of beings into
relations: these can be defined relational concepts.
Like many other natural conceptual structures relied upon in
everyday life, the distinction between classificatory and relational
concepts has been made explicit and clarified in its ontological import
by Aristotle (The Categories, 2a). Aristotle opposes such concepts as
man or horse, which classify or subsume individuals (“first substanc-
es, tode ti”), to such concepts as white or walk, which can be applied
to individuals in a relational way. The critical difference is drawn by
Aristotle himself thanks to a test based on the transitivity of defini-
tions. When a classificatory concept is predicated of an individual,
the definition of the former can also be predicated of the latter. Each
individual horse, for instance, inherits the definition of horse: if Pat
is a horse, it is “a large animal which people ride”. When a relational
concept is applied to an individual, on the contrary, the definition
of the former cannot be applied to the latter. If an individual horse
is white, it does not inherit the definition of white. A white horse is
not “the lightest colour that there is, the colour of milk and snow36”
(Collins Cobuild); it has this colour. For the same reason, an indi-
vidual galloping horse is not the action of running “very fast so that
36
See Aristotle (The Categories 2a): “suppose we take ‘white’ as an instance.
Now, ‘white’ is, no doubt, in a body and thus is affirmed of a body, for a body, of
course, is called ‘white’. The definition, however, of ‘white’ – of the colour, that
is, we call ‘white’ – can never be predicated of any such body whatever”. The
application of Aristotle’s criterion could find some difficulties in the presence
of hierarchical categories grouping central, prototypical instances and peripheral
cases. In particular, one can wonder to what extent the definition of the prototype
applies to a marginal individual instance: a penguin, for instance, does not inherit
the full definition of bird, based on the prototype (see § 4.1). The difficulty, how-
ever, bears with the substantive, empirical content of the definition, and does not
call into question its formal structure, which is based on the formal ontological
distinction between insividuals, qualities and processes. Both prototypical and non-
prototypical birds like penguins belong to the same formal ontological type, that is,
individual beings, whereas between an individual and a quality or process there is
a formal ontological gap.
122 MICHELE PRANDI
all four legs are off the ground at the same time in each stride”: it does
this action37 (Collins Cobuild).
The distinction between classificatory and relational concepts is
relevant for our topic both on substantive and formal grounds, for
conceptual consistency as well as for opposition and contradiction.
As we have remarked above (§ 1.4), the distinction plays a central role
in defining conceptual consistency and its criteria, for it provides our
shared natural ontology with its formal scaffolding. Its relevance for
opposition will be examined in the next paragraph, and its relevance
for the consistent use of contradiction in paragraph 4.
3.2.2. Closed paradigms and open series: formal lexical relations.
Thus we never see the true state of our

condition
till it is illustrated to us by its contraries,
nor know how to value what we enjoy,
but by the want of it
D. Defoe
Classificatory and relational concepts ideally belong to different

formal kinds of paradigms: whereas relational concepts tend to form
closed paradigms, classificatory concepts tend to form open series. The
distinction between closed paradigms and open series deeply affects the
balance of formal and substantive factors in determining the value of
each individual term, its dependence on the competing values and there-
fore the nature of the paradigmatic relations. Within closed paradigms,
but not within open series, it is possible to identify formal correlations
independent of the substantive contents, and therefore oppositions.
A closed paradigm imposes a system effect on its members. The value
of a term belonging to a closed paradigm critically depends on its relation-
The distinction between classificatory and relational concepts overtly involves

37
the distinction between copula – is – and other support verbs, such as have or do. The
grammar of support verbs (“verbes support”) is studied by Daladier 1978; Gross 1987;
1993; Giry-Schneider 1987. See also Vendler (1970, p. 91), who calls them light verbs.
ship with the competing values belonging to the same conceptual area,
so that formal relations are prior to individual substantive contents. As
Saussure (1916, p. 116) points out, “Within the same language, all words
used to express related ideas limit each other reciprocally; synonyms like
French redouter ‘dread’, craindre ‘fear’, and avoir peur ‘be afraid’ have
value through their opposition: if redouter did not exist, all its content
would go to its competitors”. This intuition is developed by Trier into the
idea of lexical field, which exactly depicts the essential formal property of
a closed paradigm: “The value of a word is only recognised when it is set
off against the value of neighbouring and opposing words. It has mean-
ing only as part of a whole; meaning only exists in a field” (Trier 1931, p.
45). Within a closed paradigm, lexical values are distributed among such
formal correlations as synonymy, exclusive opposition, graded opposition
or antonymy, and relational opposition, or converseness (Lyons 1963: Ch.
IV; 1977: Ch. 9). These correlations are logically prior to their substan-
tive filling, and can be defined thanks to a set of formal properties. Such
couples of values as good and bad, beautiful and ugly, or happy and sad,
for example, are all instances of one and the same formal relation, that is,
graded opposition (§ 3.3), that occurs identical independently of the dif-
ferent substantive contents that fill it in different particular cases.
When it has the structure of an open series, a field is not a network of
independent formal relations. The value of a term belonging to an open
series critically depends on the stability of its relation with the classi-
fied beings, whereas its relation with the competing values belonging to
the same conceptual area is of almost no consequence. The most typical
examples of open series are provided by the fields giving expression to
natural kinds – for instance, flowers. Unlike what happens in closed para-
digms, the gain or loss of a term in the series of flowers would not affect
the value of each individual lexeme. If a new name of flower entered the
language, or a flower lost its noun, such a circumstance would not affect
the value of such lexemes as rose, violet or hyacinth, which would go on
denoting the same natural kinds irrespective of the structural stability of
the whole field38.
38
This does not imply that an open series leaves absolutely no room for independ-
ent formal lexical structures. Such formal structures, however, do not govern the
124 MICHELE PRANDI
Within closed paradigms, formal differences are prior to individual

terms, and therefore can be identified independently of their substantive
contents. Within an open series, substantive contents are prior to differ-
ences, and the system effect is not to be perceived: a series makes visible
a set of differences only insofar as each value possesses a substantive
identity of its own. As they are grounded on substantive contents, in turn,
differences are not converted into independent formal relations. A rose is
different, say, from a violet, but this difference has no independent formal
profile, for it is entirely grounded on the substantive properties of roses
and violets. Such pairs of values as rose and violet or lily and hyacinth are
not instances of formal relation, and a fortiori of opposition.
3.2.3. Oppositions and differential paradigms.

As a formal relation independent of the organised substantive con-
tents, opposition requires the presence of a closed paradigm as a nec-
essary condition. This condition, however, is not sufficient. Our next
step will lead us to draw a further distinction between oppositions and
purely differential paradigms.
A purely differential paradigm organises a homogeneous root
meaning through a network of differential dimensions 39. The para-
digm formed by walk, run, hop, skip, crawl, for instance, organises a
conceptual space that forms a continuum, that is “the feature of move-
ment by an animate being, using the limbs” according to such differ-
ential dimensions as “the number of limbs, the order of movement, and
overall organisation of the field, but create formal islands within it. Such lexemes as
pussy and cat, for instance, form a couple of synonyms – that is, a formal relation –
included in a field that has the structure of an open series.
39
The distinction between root meaning and oppositive dimension is made by
Lounsbury 1964, pp. 1073-1074: “We shall regard as a paradigm any set of linguis-
tic forms wherein: (a) the meaning of every form has a feature in common with the
meanings of all other forms of the set, and (b) the meaning of every form differs from
that of every other form of the set by one or more additional feature. The common
feature will be said to be the ROOT MEANING of the paradigm. It defines the semantic
field which the forms of the paradigm partition. The variable features define the
OPPOSITIVE DIMENSIONS of the paradigm”.
the relation of the limbs to the supporting surface” (Nida 1975, p. 32).
The paradigm including murder, assassinate, slaughter, exterminate,
execute, slay, butcher and massacre behaves in a similar way: it organ-
ises a well delimited root meaning, coinciding with the meaning of the
hyperonym kill, owing to a network of differential dimensions. Such
paradigms do not put their values in opposition, but simply cut relevant
borders within a conceptual continuum, turning it into a network of
differences: they are differential paradigms.
An oppositive paradigm does not simply superimpose a network of
differences on a conceptual continuum, but organises a whole conceptual
area around two polar values. This is immediately visible in the presence
of binary paradigms, which contain no more than two opposite values: for
instance, alive vs dead or good vs bad40. When more than two opposite
values are available, the structure of the opposition is less apparent: two
values identify the poles, which are immediately in opposition, whereas
the remaining values grade the residual conceptual space. Within the
paradigm formed by hot, warm, lukewarm, cool, cold, icy, for instance,
warm and cold are the polar values, hot and icy intensify warm and cold
respectively, lukewarm and cool occupy the space in between.
In the presence of an opposition, no homogeneous underlying con-
tinuum, or root meaning, can be isolated independently of it. Lyons
(1963, p. 80) remarks that “the common factor, y, of good and bad is
no more easily described in terms of reference than is the meaning of
good and bad themselves”. One can certainly name the conceptual
area circumscribed by an opposition, either using a neutral term – for
instance temperature for cold vs warm – or one term of the opposition
as unmarked term: for instance, goodness for good vs bad. The avail-
ability of a label, however, does not imply that a conceptual continuum
is accessible beneath the opposition and independently of it. The oppo-
sition is logically prior.
The distinction between differential paradigms and oppositions is
well-founded in conceptual terms, but not always so easy to draw on
40
The binary structure of the paradigm does not imply that the opposition is
exclusive: see § 3.2.3.
126 MICHELE PRANDI
the field, for there are many conceptual areas where the outer form
of the two structures looks very similar and seems to merge. In such
difficult cases, the discriminating criterion is the formal or substantive
nature of the paradigm. A paradigm is formal when its form is inde-
pendent of the organised conceptual purport. It is substantive when its
form reflects the organisation of its conceptual purport. Let us exam-
ine some interesting examples.
In the area of spatial orientation, such couples as up vs down, left vs
right, forwards vs backwards are certainly in opposition. Besides shar-
ing these general oppositions, such oriented Gestalten as the human or
animal body, buildings or mountains are described by more substan-
tive couples of concepts: for instance, feet vs head, or top vs bottom.
In spite of appearances, it is not certain that such couples are really in
opposition. First, they cut relevant borders within a conceptual contin-
uum, that is, the structure of a complex object made of parts. Moreover,
the paradigm they belong to has a form that is modelled upon the form
of the object itself. This implies that there is no formal structure inde-
pendent of conceptual contents, but a network of meronymic relations
(Cruse 1986), that is, a hierarchy of substantive relations that iconically
reflect the way the complex structure is organised as a hierarchy of
parts. Among meronymic relations, there is no room for oppositions.
To take the human body as an instance, between arm and hand, leg and
foot, neck and head there is no opposition, but a substantive relation
whose form depends on the structure of the complex body itself. As
they are included in the same paradigm, such couples as head vs feet
are not formal lexical structures, and a fortiori true oppositions.
Dynamic Gestalten are described in a similar way. A stream, for
instance, has a complex structure that is characterised more by its
dynamic flowing from source to mouth than by a static form: from
space, we pass to time. Again, such couples as source and mouth
behave more like parts of a meronymic structure than like oppositions.
As Heracleitus reminds us, a river’s flowing holds as the metaphori-
cal model of the irreversible flowing of time in general, and of any
dynamic process developing in time, from a beginning to an end.
Beside such general concepts, more specific kinds of temporal proc-
esses are framed by more specific concepts: for instance dawn, sunset
or day, night as parts of the day. Again, one could be led to think that
such couples as dawn vs sunset, or day vs night form an opposition.
If we look at the whole paradigm, however, the hypothesis no longer
holds. The paradigm formed by day, night, dawn and sunset shares the
outer appearance of a graded opposition, including two polar terms and
two transition values. In spite of this, it is a differential paradigm that
cuts its formal borders within the cyclical flowing of time, which is an
experiential continuum. One supplementary argument for considering
it a differential paradigm is provided by conflict. Such an utterance as
All days are nights to see till I see thee, / And nights bright days when
dreams do show thee me (Shakespeare) looks like a contradiction but is
a shallow conflict internal to consistency, ready to be interpreted as a
metaphor, since it encourages the projection of the conceptual environ-
ment of the day on the night, and vice-versa.
3.2.4. Kinds of opposition.
Jedes ausgesprochene Wort erregt den

Gegensinn
J. W. v. Goethe
Lexical oppositions include three families carefully distinguished

by Lyons (1963, pp. 37 ff.; 1977, Ch. 9), that is, exclusive opposition,
polar, or graded, opposition, and relational opposition.
The terms of an exclusive opposition cover the whole relevant con-
ceptual area, and split it into two complementary halves, so that no
residual conceptual space is left: tertium non datur. Within the relevant
conceptual area, the terms of an exclusive opposition cannot be simul-
taneously negated, for the negation of either value coincides with the
opposite. A correlation like alive vs dead is a good instance: a person
that is not alive can only be dead; therefore, a person cannot be both
non-alive and non-dead41. The paradigms built through negation – for
41
Of course, this holds within a consistent domain: a living being cannot be both
non-alive and non-dead, whereas an inanimate being, for instance a stone, can, and
128 MICHELE PRANDI
instance, living vs non-living – necessarily form an exclusive opposi-

tion (Lyons 1977, p. 271).
A graded opposition contains at least two polar values, which
occupy two opposite points on a scale while leaving an uncovered sec-
tion, including the space in between and the extreme positions: tertium
datur. The residual sections can either be occupied by specific lexical
values or remain void, though accessible to consistent thought and
expression. Given the polar opposition warm vs cold, for instance, the
space in between hosts such values as cool and lukewarm, whereas the
extreme positions contain such terms as hot and icy. Given such cou-
ples as good vs bad or beautiful vs ugly, on the contrary, the residual
space on the scale is not occupied by specific lexical values. This
space, however, is none the less conceptually available. First, the nega-
tion of one pole does not coincide with the assertion of the opposite,
but spans the whole residual space. Therefore, the opposite poles can
be jointly negated: if a man is not good, for instance, he is not neces-
sarily bad – he can also be neither good nor bad42. Moreover, the polar
terms can be graded thanks to comparative and superlative grammati-
cal forms and modifiers: a person can be either better or less good than
another, rather good or very bad, and so on.
Relational opposition, called converseness by Lyons (1977, pp. 279-
280), involves “pairs of words which exhibit the reversal of a relationship
between items” (Palmer 1976, p. 82). Relational opposition can involve
both strictly relational and classificatory concepts. In the former case,
the same relation can be attained by two different distributions of the
same arguments. In particular, two verbs are allowed to construct the
same process on condition that the participant roles are distributed in an
actually is, but in a meta-conceptual sense – in the sense that neither alive nor dead
can be applied to it.
42
The Aristotelian tradition uses the opposition between contradictory and con-
trary for distinguishing exclusive and non-exclusive oppositions. However, this use of
the term contradiction has the disadvantage of putting on a level a virtual correlation
and an actual relation in speech. In order to avoid confusion, we shall restrict the term
contradiction to the relation in speech, and speak of opposition – either exclusive or
non-exclusive – when referring to the correlation.
opposite way among the grammatical relations: so behave, for instance,

lend and borrow. In the latter case, two opposite nouns are defined each
on the basis of its relation with the other: father, for instance, is defined by
its relation with son, and son by its relation with father. Relational oppo-
sites are logically inseparable, and this is the reason why Aristotle (The
Categories, 7b) calls them ‘correlatives’: “Correlatives are commonly held
to come into existence together […] The existence of a master involves the
existence also of a slave. If a slave exists, then must a master”, whereas
“to cancel one cancels the other43”. Thanks to their interdependence, two
converse classificatory terms, unlike the classificatory concepts belong-
ing to an open series, are involved in a formal correlation independent of
conceptual contents, and form a true opposition.
3.2.5. Polysemy, modulation, opposition and contradiction.

Most lexemes of natural languages are polysemous, that is, contain
a confederation of independent, though interconnected meanings, or
senses. Each different sense of a polysemous word undergoes different
lexical correlations, and in particular different oppositions. The adjec-
tive hard, for instance, is synonym of rigid or stiff and opposite of soft
when applied to a concrete object or material. When said of a task, it
is synonym of difficult and opposite of easy. If applied to a person and
his behaviour, it is synonym of harsh or severe and opposite of good-
hearted or sympathetic. When said of the climate, it is synonym of
bleak and opposite of mild. Polysemy bears two orders of consequences
on opposition and contradiction.
It may happen that a lexeme belongs to an opposition in one of its
senses and to a different kind of paradigm in another. Such couples as
dawn or sunrise and sunset, source or spring and mouth or outlet, for
instance, acquire a more general sense – roughly, ‘beginning’ and ‘end’
– by metaphorical extension: it is in this sense, for instance, that a man’s
fame has a dawn and a sunset. Used in such general senses, these cou-
43
Aristotle’s advice is shared by Palmer 1976, p. 82: “Lyons suggests the term
CONVERSENESS for these, but I am more concerned to point out their essentially rela-
tional characteristics, and would thus prefer RELATIONAL OPPOSITION”.
130 MICHELE PRANDI
ples are extruded from the meronymic paradigm and behave like oppo-
sitions. Accordingly, their connection forms a contradiction: Io sono il
tramonto / e l’aurora; Il tempo […] / è foce insieme e sorgente (Bracchi:
I am sunset / and sunrise; / Time […] / is both outlet and spring).
When different senses of the same lexeme belong to different oppo-
sitions, contradiction does not arise if more than one relevant sense is
involved – if one sense is connected to the opposite of another. When
Midas is said to be divesque miserque (Ovid), or Eugénie Grandet is
qualified as la pauvre riche héritière (Balzac: the poor rich heiress),
poor is not taken as the opposite of rich but of happy. As their burning is
metaphorical, the eyes are in fact flameless: Faci, che sempre ardete, e
siete spente (Melosio: Stars ever flaring, and yet spent). Hegel’s shocking
statement that Das Böse […] ist die positive Negativität (Evil is positive
negativity) is no contradiction because the connected concepts are not in
opposition. The term negativity refers to the fact that evil is the negative
pole of the opposition whose positive term is good. The term positive
means that evil is not just the absence of good but its antagonist force,
which shares with it the positive reality of possessing an active power44.
As Cruse (1986, p. 52) points out, a single sense of a lexeme, inde-
pendently of polysemy, “can be modified in an unlimited number of
ways by different contexts, each context emphasizing certain semantic
traits, and obscuring or suppressing others […] This effect of a context
on a lexical unit will be termed ‘modulation’”. One effect of modula-
tion in the presence of relational concepts is to create local oppositions
between values that are not directly in opposition from an immanent
lexical point of view. The opposite of happy, for instance, is sad, and
44
Owing to its aptitude to combine consistent thought and the rhetoric appeal
of contradiction, the apparent oxymoron is praised by Hegel, who throws it against
the identity principle as a true war-machine. It is interesting to compare Hegel’s way
of forcing a conflict into logical contradiction to the analytical attitude displayed by
Kant 1763) when facing a similar topic: “Vice (demeritum) is not merely a negation; it
is a negative virtue (meritum negativum). For vice can only occur in so far as a being
has within him an inner law […] which is contravened by his actions. This inner law
is a positive reason for a good action […]. What we have here is, accordingly, a dep-
rivation, a real opposition, and not merely a lack”.
not frightened. In spite of this, when Tolstoy writes that Anna was both
frightened and made happy thereby one indisputably perceives an oxy-
moron. The reason is that in this co-text fear is dramatically focalised
as the antagonist of happiness. Instead of resting on an independent
lexical opposition, this expression treats a conflict as if it were a con-
tradiction, forcing its terms into an opposition.
4. A typology of consistent contradictions.
Es hielo abrasador, es fuego helado,

es herida que duele y no se siente,
es un soñado bien, un mal presente,
es un breve descanso muy cansado.
Es un descuido que nos da cuidado,
un cobarde con nombre de valiente,
un andar solitario entre la gente,
un amar solamente ser amado.
Es una libertad encarcelada,
que dura hasta el postrero parasismo,
enfermedad que crece si es curada.
Éste es el niño Amor, éste es su abismo.
¡Mirad cuál amistad tendrá con nada
el que en todo es contrario de sí mismo
F. de Quevedo
All the distinctions outlined above are relevant for our topic. Our
next step will be to check, for all kinds of form of expression and for
all kinds of concept and correlation, which distributions end in con-
tradiction and which do not, and which contradictory expressions are
ready to convey consistent concepts, on what conditions and with what
outcomes. In the following sections, we shall examine first the distri-
bution of classificatory concepts, and then the distribution of relational
concepts.
4.1. Classificatory concepts.

When two classificatory concepts belonging to the same series
are distributed within a subject-predicate structure – This square is a
triangle; This dog is a cat; This surgeon is a butcher – the outcome
132 MICHELE PRANDI
is not contradiction, but a conflictual classification. This conflict is

consistent because it does not challenge the ontological identity of the
subject – for instance the identity of the surgeon as a human being,
or of the dog as an animate, non-human being – but only its empiri-
cal identity – the identity of the surgeon as a surgeon, or of the dog
as a dog. Clearly, the empirical identity as circumscribed by different
kinds of classificatory concepts is a layered property, which can be
affected more or less deeply by conflictual classification. The identity
of a dog as a dog, that is, as a member of a natural kind, is more deep
and vulnerable than the identity of a surgeon as a surgeon. Moreover,
there are beings whose identity is not empirical at all. The identity of
a geometrical figure, for instance, is neither ontological nor empirical
but analytical, that is, tautological.
A conflict at empirical level provides the conceptual ground for
metaphorical interaction. In metaphor, challenging the identity of a
being is a preliminary condition for imposing a strange model on it:
the surgeon, for instance, is viewed through the model of the butcher;
a dog has feline characteristics (Haser 2005, p. 29). When analytical
concepts are in conflict – This square is a triangle – no interaction is
at hand. The conflict is no more than a cramp of thought.
When two concepts belonging to the same series are jointly applied
to a subject –This figure is a square and a triangle; This building is a
cottage and a castle – the outcome is a conflicting classification that
challenges the identity of the subject but is barred to metaphorical
interpretation. In the presence of analytical concepts, we are simply
faced with a cramp of thought: one cannot even imagine in what sense
a figure can be both a square and a triangle. In the presence of empirical
concepts, one can always figure out some bizarre entity: for instance, a
professional figure part surgeon and part butcher, a building part castle
and part cottage, a mythological being like a centaur or a chimera. Or
one can think of some of Magritte’s paintings, for instance L’Explication
(1954), which blends a bottle and a carrot to form a sort of missile45.
45
As Cardinal 1974 points out, in Magritte’s L’Explication “we are shown first a
bottle, then a carrot, and finally a ‘carrottle’ ”. As the quotation shows, this form of
In the realm of empirical concepts, conflictual classification is

compatible with consistent thought, owing to the complexity of things,
which escapes rigid classification; in the realm of analytical concepts,
it is not, for the individual identity has no conceptual purport beyond
a purely tautological set of formal necessary and sufficient conditions.
Such expressions as This surgeon is a butcher, or even This building
is a cottage and a castle, can certainly depict things in a relevant and
perspicuous way. Such expressions as This square is a triangle or This
figure is a square and a triangle as certainly cannot.
Unlike the correlation between two classificatory concepts, the
correlation between a classificatory concept and its negation forms an
opposition, so that the joint application of the two terms to the same
being ends in contradiction: This cottage is not a cottage; This building
is and is not a cottage. In the realm of empirical concepts, this form of
contradiction is open to consistent thought, and the conceptual basis
for this lies in a property of natural taxonomic categories first focused
on by Black (1952) and then widely studied by cognitive psycholo-
gists and linguists. Natural categories do not follow the Aristotelian
principle (Categories, 3b) according to which “no substance can admit
of degrees in itself46”. On the contrary, our experience proves that an
individual referent may satisfy the requirements for its inclusion in a
given category to a variable extent. As Black (1952, p. 28) points out,
“If we examine instances of the application of any biological term, we
shall find ranges, not classes – specimens arranged according to the
degree of their variation from certain typical or ‘clear’ cases”, that
is, from a reference point or model reminding of Weber’s idealtypus
(Weber 1922), and later called prototype Rosch (1972; 1973; 1975;
double classification is lexically productive in the creation of endocentric portmante-

au words such as smog or brunch. This structure is not shared by exocentric tokens
such as motel, which condense a subordinative structure head-modifier: whereas
smog is both smoke and fog, a motel is not both a motor and a hotel, but a hotel located
along a motorway.
46
See also Categories, 3b: “For example, the same substance, man, cannot really
be more or less man as compared with himself or another […] Substance can have
no degrees”.
134 MICHELE PRANDI
1978; Taylor 1989; Kleiber 1990). A prototype is neither an individual

paradigmatic instance47 – because its range of application is extended
to instances that do not match a given image – nor a simple collec-
tion of empirical data – because it has a hierarchical structure which
imposes a strong perspective on experience data. A prototype is not
part of actual experience, but a model filtering our access to experi-
ence48. The prototype of bird, for instance, is neither the image of a
given bird nor a collection of random data about birds, but a hierarchy
of shared ideas about what a bird should ideally be and what a bird
is empirically allowed to be – a filter organising the experience of
real, empirical birds. This implies that some properties are constitu-
tive of the category without being either necessarily shared or shared
47
As Taylor (1989, p. 59) points out, “There are two ways in which to understand
the term ‘prototype’. We can apply the term to the central member, or maybe to the
cluster of central members, of a category. Thus, one could refer to a particular artefact
as the prototype of CUP. Alternatively, the prototype can be understood as a schematic
representation of the conceptual core of a category. On this approach, we would say
not that a particular entity is the prototype, but that it instantiates the prototype”.
The second alternative sharply dissociates concept and image, which confers it an
undisputable Kantian flavour: see Kant 1781, pp. 182-183: “Indeed it is schemata, not
images of objects, which underlie our pure sensible concepts. No image could ever be
adequate to the concept of a triangle in general. It would never attain that universality
of the concept which renders it valid of all triangles, whether right-angled, obtuse-
angled or acute-angled; it would always be limited to a part only of this sphere […]
Still less is an object of experience or its image ever adequate to the empirical con-
cept […] The concept ‘dog’ signifies a rule according to which my imagination can
delineate the figure of a four-footed animal in a general manner, without limitation
to any single determinate figure such as experience, or any possible image that I can
represent in concreto, actually presents”. Kleiber (1990, p. 158) discusses a connected
point, that is, the relationship between prototype and Wittgenstein’s idea of a network
of family resemblances, a polycentric cognitive structure that leaves no room for a
central type. On the difference between monocentric and polycentric categories, see
also Geeraerts 1988.
48
According to D’Andrade (1987, p. 112), “A cultural model is a cognitive schema
that is intersubjectively shared by a social group”. Insofar as it holds as a schema, a
model is not part of actual experience, for “experience does not direct us to derive
anything from experience” (Wittgenstein 1969, prop. 130); insofar as it is intersubjec-
tively shared, a model is not a psychological content.
to the same degree by all the instances that are subsumed under the
category. The ability to fly, for instance, is an essential property of
the prototypical bird but is not shared by all kinds of bird. Conversely,
instances which do not share the typical properties are admitted
within the category with a peripheral status, which do not affect the
content of the category. Birds that cannot fly are called birds but do
not contribute to the prototype.
Owing to both the hierarchical structure of classifying categories
and the complexity of empirical data, any instance to be found in
actual experience has to negotiate the degree to which it satisfies the
idealtypus. Accordingly, the use of classificatory concepts is insepa-
rable from judgement, evaluation and decision, and the awareness of
this emerges from both linguistic coding and use. Many lexical hedges
such as real or a sort of are meta-conceptual devices, so to speak,
grading the commitment on the part of the speaker to a given act of
categorisation. A real brother, for instance, is a person who fully meets
the relevant criteria for brotherhood, whilst a sort of jacket is meant
to be a jacket in a loose sense only. Classificatory nouns can even co-
occur with grading adverbs: The grey of Pablo […] is much horse”
(Hemingway). Along with tautology, and playing an opposite function,
contradiction is part of this repository of means. A tautology like This
woman is a woman can be used to suggest that an instance is taken as
a good example of the type49, whereas contradiction calls into ques-
tion this match. The subject-predicate form, which directly negates
the identity of the subject – This woman is not a woman – underlines
that the referent is not worthy of the concept, whereas the coordina-
tive form, which balances the opposite concepts – Ann is and is not
a woman – stresses that the subject does not fully match the require-
ments of the concept50. In the same way, birds that cannot fly, such as
49
In Italian, reduplication, which is similar to tautology, is sometimes used with
this value: a caffè caffè (“coffee coffee”) is a kind of coffee which “instantiates the
typical qualities […] of real coffee” (Grandi 2002, p. 256).
50
The interpretation of the last form of contradiction makes visible an interest-
ing property of the textual use of classificatory concepts. When interpreting such an
utterance as Ann is and is not a woman, one takes the two occurrences of the noun
136 MICHELE PRANDI
penguins, are and are not birds. When describing with telling irony a
cottage that is not really a cottage, Jane Austen seems to have just read
a paper by Eleanor Rosch:
As a house, Barton Cottage, though small, was comfortable and compact; but
as a cottage it was defective, for the building was regular, the roof was tiled,
the window shutters were not painted green, nor were the walls covered with
honeysuckles.
4.2 Classifying through relations: relational opposition.
Nec refert, dominos illic famulosne requiras:

Tota domus duo sunt, idem parentque iubentque
Ovid
Classificatory concepts involved in relational oppositions, called

‘correlatives’ by Aristotle, are non-saturated concepts, which do not
classify beings thanks to the inherent properties of a kind taken in
isolation, but owing to a relation involving a couple. A father is not a
woman as if they conveyed two distinct conceptual perspectives upon the woman –
two different focalisations or modulations (Cruse 1986, p. 52) imposed on the set of
relevant features constitutive of the concept: for instance, the woman in a physical and
in a moral sense. The use of a given concept in connection with a given entity does not
require the relevance of all its typical features to the particular instance. The fact that
an entity shares one constitutive property of a concept is considered by the language
user a sufficient condition for applying the concept to it. As a consequence, the dif-
ferent textual occurrences of a word do not necessarily convey isomorphic instances
of the same general concept. If the utterance Ann is a woman is used to suggest that
Ann has reached physical maturity, for instance, the speaker does not necessarily
commit himself to the whole set of properties that qualify the prototypical concept of
womanhood – for instance, to psychological maturity. Accordingly, the sequence Ann
is a woman, but you wouldn’t entrust her with a child is not contradictory. In other
conditions, a speaker can use the same utterance in order to focus on psychological
maturity, without committing himself to physical maturity. Accordingly, a sequence
like Ann looks rather childish but she is a woman is not contradictory. It is on this
background that a contradictory expression like Ann is and is not a woman can sug-
gest that Ann satisfies only part of the set of properties constitutive of prototypical
womanhood.
father in the way a dog is a dog; a slave is not a slave in the way a cat
is a cat. A father is a father because he has a son, and a son is a son
because he has a father. A slave is a slave because he has a master,
and a master is a master because he has a slave. In the presence of
classificatory concepts, each kind of being enters into an open series
provided with an independent identity of its own. In the presence
of converse concepts, the value of each term depends on the formal
relation with the correlative value within a closed paradigm. The
classification through relation is eccentric with regard to the inherent
classification of natural kinds, and presupposes it. Accordingly, the
purely classificatory identity of a being, both ontological and empiri-
cal – the identity of a cat as a cat, or of a person as a person – is
not affected by the imposition of a relational identity such as father
or son, or slave or master: a son is no less a living being – a cat or
a person – than a father; a slave is no less a person than a master.
This is the reason why a conflictual classification of a being affects
its identity as an individual at a more or less deep level – This cat
is a dog; My brother is a lion – whereas the joint attribution of two
converse concepts to the same subject – John is father and son; John
is slave and master – does not. As the last examples show, however,
one thing is the identity of the subject, another is the compatibility
of the converse concepts engaged in shaping the content of the predi-
cate. In order to feed consistent thought, both conditions have to be
satisfied.
When relational oppositions circumscribe the changing processes
of social and cultural life, contradiction is ready to make room for
consistent thought. A typical case, which runs along the history of
Western philosophy from Aristotle to Hegel and Marx, is the couple
slave vs master. The relationship between a slave and a master is com-
plex, potentially conflictual and dynamic. It is complex because it is
gradable: while the slave depends on the master for power and right,
for instance, the master may depend on the slave for sustenance, and
therefore for life itself. Awareness of this fact may rouse a conflict,
which is liable to overturn the relation: the same person can pass from
the condition of slave to the condition of master and vice-versa without
affecting his identity as a person.
138 MICHELE PRANDI
When relational oppositions are rooted in the rigid constraints of

nature, contradiction is no longer a useful tool for thinking. The diffi-
culty, however, does not stem from the contradictory expression itself,
but from the hard structure of things, where real incompatibilities leave
no room for real oppositions. One example is the behaviour of rela-
tional opposites belonging to kinship paradigms. If the same person
is said to be both mother and daughter, for instance, her identity as a
human person is not called into question, for either opposite predicate
is consistent with the subject. This, however, does not hinder the hard
fact that the same person cannot be both mother and daughter of the
same individual, who in turn would be both her father and son. This
point is illustrated ex negativo by the mystery of the Virgin, who is
both God’s mother and daughter: Vergine Madre, figlia del tuo figlio51
(Dante: Virgin Mother, daughter of thy son52).
The couple of personal pronouns I and you53 is a peculiar case of
relational opposition. These pronouns form a relational opposition
insofar as they denote the changing roles and turns in dialogue: I am
I, the person who plays the role of speaker, as long as you remain
you, and I am ready to take up the role of you as soon as you become
I. Beside this purely relational use, however, the personal pronoun I
51
Kinship terms denote two-places relationships: an individual x is either father,
or mother, or brother, and so on, with regards to the same individual, y. Kinship
paradigms are complex enough to combine two different kinds of correlation, that is
relational oppositions – for instance, father vs son; mother vs daughter – and differ-
ential correlations: for instance, father, mother, brother. If two relational opposites are
jointly predicated of one and the same couple of individuals, there is contradiction, as
in Dante’s quoted verse. If different terms of a differential paradigm are jointly predi-
cated of one and the same couple of individuals, there is shallow conceptual conflict,
liable to be interpreted as metaphor. In Andromache’s words, the typical qualities of
a father, of a mother, of a brother and of a husband are jointly projected on Hector:
Hector, you are to me father and queenly mother, you are brother, and you are my
vigorous husband (Murray 1999, Il. 6).
52
Amati Parker 2006.
53
According to Benveniste 1956, only I and you are personal pronouns in a nar-
row sense, that is, expressions of the correlative roles (dramatis personae) in com-
munication, whereas so called third person forms are ready to denote anything else
one can speak about.
is highlighted by its function of rampart of personal identity. Every

time one says I, one tacitly presupposes one’s identity: I am what I
am. Unlike the link with the role of speaker, which is grounded in the
meaning of the pronoun I, the link between I and the personal identity
of an empirical individual is an ephemeral indexical link. Moreover,
the set of properties that form the ground of personal identity is cer-
tainly taken for granted by the individual person who says I, but this
implies neither that these properties form a consistent network, nor
that the subject is really aware of them. One thing is to say I, another
is to know oneself – gnozein seauton, to borrow the Oracle’s recom-
mendation made famous by Socrates. The contradictory expression I
am not what I am (Shakespeare) theatrically shows how sliding can be
the very ground of personal identity one takes for granted each time
one says I54.
For the same reasons, personal identity sheltered by the pronoun
I is not directly opposed to the addressee’s identity but simply differ-
entiated from the identity of any other self, including the occasional
addressee. As an index of an individual self, I no longer forms a rela-
tional opposition with you, but belongs to an open series of referring
expressions – pronouns, proper nouns and noun phrases – each capa-
ble of denoting in each situation of use an individual self belonging to
an immense open series of individual selves to be found in the past,
present and future world. When Maria says to Robert Jordan I am thee
and thou art me and all of one is the other (Hemingway), there is no
contradiction, but an impulse of fusion between two distinct individu-
als. The pronoun thou is no more than the contingent index of another
self, just as the noun Silvia in Shakespeare’s lines: To die is to be
54
On the one hand, Benveniste 1956 reminds us that I and you are simply the
mask one puts on when assuming one’s role in the comedy of speech. On the other
hand , each individual fills the void role with a specific content, which is gathered
under the label of self-awareness. Within the philosophical tradition opened by
Descartes’ Cogito, ergo sum and including Kant, Fichte and Husserl, self-awareness
summarised by I is looked at as the very ground of everything, from episthemology
to moral sense and aesthetics. How fragile a foundation, indeed.
140 MICHELE PRANDI
banish’d from myself; / And Silvia is myself: banish’d from her / Is self
from self: a deadly banishment!
4.3. Relational concepts.
Né pentere e volere insieme puossi

per la contradizion che nol consente
Dante Alighieri
The behaviour of relational oppositions, which is eccentric among

classificatory concepts, is due to the relational and formal component
of the paradigm. The relational component of classification displaces
the focus from the inherent identity of the subject – a dog is a dog, a
person is a person – towards a chain of relational determinations – a
complex condition or a story – that presupposes it: a person that is part
slave and part master, or overturns his condition from slave into mas-
ter, remains a person. This dynamic relationship between complexity,
conflict, change and firm identity is precisely what characterises the
application of relational concepts to beings.
A being cannot escape its natural kind. A person, for instance, is
not destined to become a dog or a tree, as happens among others to
Daphne in Ovid’s Metamorphosis: Mollia cinguntur tenui praecordia
libro, / in frondem crines, in ramos bracchia crescunt; / pes modo tam
velox pigris radicibus haeret, / ora cacumen habet; remanet nitor unus
in illa. But the same person can pass from youth to old age, from hun-
ger to satiety, from sleep to waking or from love to hate without losing
his identity as a person. It is owing to this that the synthetic predication
that jointly attributes two opposite and consistent relational predicates
to the same subject is the elective form of contradiction for depicting
complex, conflicting and changing situations. Two opposite properties
or processes coexist in a complex situation, ready both to conflict with
and to turn into one another. The use of this form of contradiction as
an instrument for consistent thought is typically documented in the
area of human emotions, where competing opposite forces of grada-
ble intensity are likely to attain different points of balance changing
in time: Once again I love and I do not love, / I am mad and I am
not mad (Anacreon55); Odi et amo (Catullus); Ardo et son di ghiaccio

(Petrarch); I both wished and feared to see Mr Rochester (Charlotte
Brontë); Vorrei e non vorrei (Da Ponte), She was both frightened and
made happy thereby (Tolstoy). Opposite human emotions form graded
oppositions, so that the negation of one pole does not coincide with the
assertion of the opposite, but covers the whole residual space, includ-
ing the bare absence. Whereas Catullus’ I hate and love describes a
conflict between love and its antagonist, Anacreon’s I love and I do not
love opposes love and its absence.
A relational term denoting a process associates two functions: it
classifies instances under categories – singing or smiling, for instance
– and relates referents: for instance, it attributes the process of smil-
ing or singing to a human being. A process is typically expressed by
a verb occurring in a finite form, which occupies the “gravitational
centre of the sentence” (Humboldt 1836), surrounded by its arguments.
However, it is also accessible to both a noun and a verbal infinitive or
gerund, which are the noun-like forms of the verb. Owing to its gram-
matical properties, the finite verbal expression makes immediately
visible the relational side of the concept; its contradictory use is well
documented by the above examples. A nominal form, on the contrary,
highlights the classificatory side of relational concepts – namely the
function of gathering processes in consistent classes.
A noun, an infinitive and a gerund can occur both as subject and
as predicate: The battle was cruel; This is a battle. In both forms, the
noun classifies instances. The predicative structure has the form of
a definition, which gives explicit expression to the same act of clas-
sification that remains implicit in the referential form. Different kinds
of conflicting classification of processes, with different formal and
conceptual properties, can be grafted onto predication owing to the
nominal distribution.
A process and its opposite can jointly be applied to the same sub-
ject, which is classified in a contradictory way: This is and is not a
battle; This feeling is both hate and love. If two predicates belonging
55
Campbell 1988.
142 MICHELE PRANDI
to a differential paradigm are jointly attributed to the same subject,

they cumulate without conflict: Ann is singing and speaking; John is
hopping and running.
Within the subject-predicate frame, as we have already remarked,
contradiction through negation is the negation of a tautology. Such a
contradiction as Love is not love explicitly negates the tautology Love
is love: Love is not love / Which alters when it alteration finds, / Or
bends with the remover to remove (Shakespeare).
As the contradictory classification of a thing, the contradictory
classification of a state of affairs through negation – This is and is not
a battle; This battle is not a battle – calls into question its empirical
identity. In the former example, a given event is not considered a real
battle; in the latter, an event called battle does not deserve the name –
and one immediately thinks of the conversation between Fabrizio del
Dongo, the hero of Stendhal’s La chartreuse de Parme, and a soldier
on the battlefield of Waterloo: – Mais ceci est-il une véritable bataille?
– Un peu (But is this a real battle? – Sort of ).
In the area of lexical paradigms, the relevant difference is once
more that between oppositions – for instance Hate vs love, cry vs
laugh – and differential paradigms: for instance walk, run, hop,
skip and crawl. In the presence of opposition, identification ends in
contradiction: for instance His love is hate; To give away yourself
keeps yourself still (Shakespeare). A contradiction of this form – for
instance, Love is hate – negates the correlative tautology – Love is
love – implicitly, owing to the form of the lexical opposition. In the
presence of a differential paradigm – Hamlet’s walking is crawling,
Ann’s speaking is singing – the outcome is not contradiction but a shal-
low kind of conceptual conflict, which challenges empirical identity
within the borders of consistency, bound to end in metaphor: Ann’s
speaking sounds as a kind of singing; Hamlet sees his walking as if it
were a kind of crawling: What should such fellows as I do crawling /
between earth and heaven?
Properties are typically entrusted to a dedicated category, that is,
adjectives, but can also be expressed by nouns. Just as in the case of
processes, the nominal expression highlights the classificatory side of
properties, and in particular makes both tautological and contradictory
forms of classification possible: for instance, This is and is not joy;

Sadness is sadness; Ann’s joy is sadness; Ann’s joy is no joy; Ann’s joy
is sad.
When the opposite properties are expressed by adjectives, it is
worth comparing the behaviour of graded and exclusive oppositions.
The polar terms of graded oppositions, defined contraries by
Aristotle, leave a residual conceptual space uncovered, to which neither
fully applies. In purely conceptual terms, this space admits determina-
tion both in negative terms, stressing the unfitness of either opposite
concept – John is neither good nor bad – and in positive terms, as a
variable mix of both: John is good and bad. The double negation is logi-
cally consistent, whereas the double assertion ends in contradiction.
In spite of the logical advantage of double negation, it is interesting
to note that the same speakers who do not hesitate to use a contradiction
to express a complex or conflicting situation are reluctant to depict it as
the bare absence of determinations. As Sapir (1944, p. 101) points out,
“To the naive, every person is either good or bad; if he cannot be easily
placed, he is rather part good and part bad than just humanly normal or
neither good nor bad”. The average state between the opposite properties
tends to be hit by a sort of moral censure, an attitude that has become
proverbial in Dante’s allusion to l’anime triste di coloro / che visser
sanza ‘nfamia e sanza lodo (the melancholy souls of those / Who lived
withouten infamy or [lit. and] praise56). Instances of it are to be found
both in everyday discourse – for instance in the saying To be neither
fish, nor fowl, nor good red herring – and in narrative: Within the coach
was sitting a person who, though not being very handsome, looked not
so bad, neither too fat not too slim; one could not tell that he was old,
but he wasn’t too young either (Gogol); It was not only that I could not
become spiteful, I did not know how to become anything; neither spite-
ful nor kind, neither a rascal nor an honest man, neither a hero nor an
insect (Dostoevskij57); [He] now wandered in a kind of limbo, because
56
Just as proverbial is Virgil’s commentary: Non ragioniam di lor, ma guarda e
passa (Let us not speak of them, but look, and pass, Amari Parker 2006).
57
Engl. transl. by Constance Garnett, Electronic Text Center, University of
Virginia Library, http://etext.lib.virginia.edu/modeng/modengD.browse.html.
144 MICHELE PRANDI
he wasn’t good or bad enough […] Life didn’t exist any more (Greene).
Whereas the absence of both opposite qualities is hit by a sort of cen-
sure, their balance is praised from within an ethics and poetics inspired
by the maxim of aurea mediocritas against the extreme poles58: “In one
of Bandello’s sonnets, for instance, the mix [mistura] of contraries is
seen as safeness against both absolute positivity and absolute negativ-
ity: Così mi regge Amor, che s’a quest’alma / desse solo martir o gioia
pura, / col peso ne morrei di tanta salma. / Ma mentre l’un con l’altro
fa mistura, / morte non può di me portar la palma, / ché se m’impiaga
l’un, l’altro mi cura (XXXI, 9-14: In such a way love affects me, if it
bestowed upon this soul only pain or pure joy, with the weight I would
die. Yet as one is mixed with the other, death cannot claim me, since one
wounds while the other heals)” (Gigliucci 2004, p. 17).
Exclusive oppositions leave no residual conceptual space, so that
no empirical case can escape the sharp alternative: a living being, for
instance can only be either alive or dead. The absence of a residual
space for grading seems likely to deprive contradiction of any consist-
ent conceptual content. In fact, the logical impossibility of a residual
conceptual space is counterbalanced by the presence of a grey zone
that is empirically indeterminate between the opposite alternatives.
There are cases when it can be very difficult to decide whether a given
being satisfies a predicate or the opposite, and even more to decide
whether a being is or is not included within the consistent range of
application of an opposition, and yet the decision cannot be avoided
on moral grounds. For instance, it is very hard to identify the exact
moment when a person passes from life to death, and even more to
state when an embryo becomes a living person, open towards life and
death. Yet, there are circumstances where the decision, however hard
on empirical grounds, cannot be avoided in practical behaviour.
58
This wavering between despising what is just ‘humanly normal’ and celebrat-
ing aurea mediocritas is just one case of a more general tendency of common wisdom
to adhere at different moments and with the same force to opposite attitudes. Among
proverbs, for instance, almost every statement is counterbalanced by its opposite: on
the one hand, Clothes do not make the man; on the other hand, The tailor makes the
man (Hamm 1989; see also Kleiber1998).
Qualities are the very realm of oppositions, both exclusive and

graded, but do not ignore the differential form of organisation, as is
shown by the field of colours. However identified and differentiated
thanks to external cognitive anchors in a way that reminds one of natu-
ral kinds, colour lexemes do not form an open series but a hierarchy
of closed paradigms, so that the value of each lexeme depends on the
form of the whole field59. These paradigms, however, are purely dif-
ferential. Therefore, the joint application of two different colour terms
to the same subject ends neither in contradiction, as in the presence of
opposition, neither in shallow conflict, as in the presence of different
classifications of the same being, but simply cumulates the two prop-
erties: Ann’s skirt is yellow and red plainly describes a two-coloured
surface. The same being that cannot jointly receive opposite properties
without contradiction – Ann is happy and unhappy – can jointly receive
different properties. The identification of different colours, for its part,
ends in shallow conflict, as it is the rule in the presence of differential
paradigms: rubor negro60 (García Lorca: black blush).
When used in an extended sense, some colour lexemes are taken
into oppositions, and therefore are open towards contradiction. In
English, for instance, green is the opposite of ripe when predicated
of fruits. In Spanish, like in Italian and French, verde, green, is the
opposite of seco, dry, when predicated of trees: Arbolé arbolé seco y
verde (García Lorca).
59
Though language-specific, even highly idiosyncratic for their form, colour con-
cepts depend on a shared experience for their content. The differentiation of colour
concepts does not rely on independent formal relations internal to the paradigm, but
rests, just as in the area of natural kinds, on the identification of salient differences
to be perceived within a largely shared, if not universal experience – the experience
of seeing things against familiar backgrounds: for instance, “the sky (often blue), the
ground (often brown), the grass (typically green), the sun (often yellow and brilliant),
the sea (often dark blue), the broad expanse of snow (normally white)” (Wierzbicka
1996, p. 289).
60
There is no agreement among experts about this passage: according to some
criticians, the correct spelling is rumor negro (black noise), which would be a syn-
esthesia.
146 MICHELE PRANDI
Adjectives can occur both as modifiers of a nominal head – a happy

girl – and as predicates applied to a subject: Ann is happy. Both forms
are open towards contradiction: I lived six years in this happy but
unhappy condition (Defoe); This girl is happy and unhappy. The head-
modifier relation, for its part, is shared by adverbs modifying verbs:
John walked slowly. Contradiction can graft on modification in two
ways. Either two modifiers are applied to the same head – John walked
slowly and fast; I lived six years in this happy but unhappy condition
(Defoe) – or the modifier contradicts the head: ‘Twas not my blame –
who sped too slow (Dickinson); It always seemed to me a sort of clever
stupidity only to have one sort of talent (George Eliot).
The most interesting property of modification is its structural
instability, which can be observed under ideal conditions in the case
of adjectives thanks to the comparison with the alternative use in
predication. The two uses open to adjectives are not equally fit for their
elective function, that is, for the attribution of a property to a being, or
qualification.
Modification acts at noun phrase level: this red apple. As it is not
supported by an independent grammatical relation, the head-modifier
link behaves as a weak mould, ready to negotiate the connection with
the connected concepts (Prandi 2004, pp. 130-144). This implies that
within a noun phrase one can find both adjective that do not ascribe
qualities to the referent of the head noun and adjectives that attribute a
quality outside the head-modifier link. The former case is documented
by relational adjectives61. When applied to a head noun, a relational
adjective does not ascribe a quality to its designatum, but connects
it with another entity. The nervous system, for instance, is a system
61
Bally (1932, p. 97) describes the most relevant properties of relational adjec-
tives in French. If we leave aside position, his remarks also hold for English: “Thus in
chaleur solaire, solaire cannot come before the noun (solaire chaleur being impos-
sible); it cannot be modified by adverbs normally used with adjectives, so that chaleur
très solaire cannot be said; last but not least, it cannot be a predicate, so that Cette
chaleur est solaire would be unintelligible”. It is mainly owing to the presence of
relational adjectives that a consistent definition of the category adjective must address
the question of typicality: cf. Dixon 1977; 1994; Bhat 1994.
made of nerves. When the head noun is a noun of process, a rela-

tional adjective is ready to express one of its arguments or margins.
The expression The Indian war may refer to a war involving Indians
or fight in India. A monosyllabic preoccupation (Hardy) is a preoc-
cupation expressed in monosyllables. When the head noun denotes
a property, the adjective expresses its bearer, reversing the canonical
distribution, as in rural wilderness. The latter case is documented by
oblique modification. Under the pressure of conflicting concepts, the
relation of qualification can be displaced outside the head-modifier
link to involve an entity somehow connected with the head. In its most
natural interpretation, the expression blonde ambition (The Guardian)
does not attribute blonde hair to ambition in a metaphorical way, but to
an ambitious person in an oblique way; the expression white wonder of
dear Juliet’s hand (Shakespeare), attributes whiteness to Juliet’s hand
passing through another quality, that is, wonder.
Predication acts at sentence level: This apple is red. The core of a
sentence is a network of empty grammatical relations, insensitive to
the connected conceptual contents. Owing to this, it provides qualifi-
cation with a strong formal mould. Unlike a noun phrase, the predica-
tive structure filters out all the adjectives and uses of adjectives whose
function is not to attribute a property to the referent of the subject.
These non-canonical uses of adjective are all incompatible with the
rigid formal scaffolding of sentence structure: sentences like The poli-
ceman is rural, or This preoccupation is monosyllabic sound at least
unnatural62, for the predicate can hardly be applied to the subject as a
true quality.
On this account, it is no wonder that the weak head-modifier link
is the favourite formal tool for entrusting both personal disagreements
and ideal conflicts to contradiction: It always seemed to me a sort of
62
According to Bolinger (1967, p. 15), they are even ungrammatical: “The
ungrammaticality of *The policeman is rural illustrates the divergent restrictions that
apply to the two uses of the noun, as subject of a predication and as part of a noun
phrase”. Predicative use is allowed on condition that the adjective loses its relational
character and is interpreted as qualifying, in general by analogy, as for instance This
day is wintry.
148 MICHELE PRANDI
clever stupidity only to have one sort of talent (George Eliot); com-
monplace occupations, which amuse the mind with facile difficulties
(Flaubert); I shall have accomplished a blessed crime (Sophocles);
Living death (Campanella). This preference documented in texts, how-
ever, does not imply that such a dissociation of responsible subjects is
incompatible with predication: Jugum meum suave est, et onus meum
leve (Matthew).
5. Concluding remarks.
Cumque sit ignis aquae pugnax, vapor umidus omnes

res creat, et discors concordia fetibus apta est
Ovid
This paper has been aimed at defining under what formal and con-
ceptual conditions a contradiction is capable of expressing a consist-
ent thought. In general terms, this is possible because contradiction
and conceptual consistency rest on independent orders of lawfulness.
Contradiction depends on opposition, which is a formal structure,
either syntactic or lexical, whereas conceptual consistency rests on
the substantive content of a shared natural ontology. The expression
of complex, conflicting and changing situations typically requires
the co-occurrence of couples of opposite concepts – for instance, love
and hate – which can be framed in a contradictory expression, as well
as in a non-contradictory one. Insofar as the opposite concepts are
equally consistent with the subject, no obstacle is opposed to consistent
thought in either case.
Turning from conceptual to interpersonal level, a contradiction is
ready to express a conflict between persons and Weltanschauungen.
In such cases, the speaker subscribes to one of the opposite concepts
while disowning the other, whose responsibility falls upon another
subject. The common moral sense, for instance, may approve as a
blessed deed what the tyrant’s law condemns as a crime, as happens
with Antigones’ blessed crime.
The use of contradiction to frame consistent thought is subject to
a set of conditions depending on the structure of the contradictory
expression, on the form of the paradigms involved and on the nature of

the concepts opposed.
Contradiction takes two main forms located at sentence level, that
is, the joint application of two opposite predicates to the same subject
– for instance I both wished and feared to see Mr Rochester – and
the negation of a tautology: for instance, This battle is not a battle.
Whereas the first form is a kind of synthetic predication, which is the
elective form of expression of consistent thought, the second somehow
challenges the identity of the subject. A third form is the head-modifier
relation, combining a noun and an adjective – a sort of clever stupidity
– or a verb and an adverb: ‘Twas not my blame – who sped too slow. In
all these cases, the involved concepts are the terms of an opposition,
which is a formal kind of relation.
A concept and its negation necessarily form an opposition, no matter
if the concept is classificatory or relational, member of an open series
or of a differential paradigm. Turning to lexical relations, opposition
presupposes the accessibility of a formal kind of lexical relation. Two
interconnected relevant distinctions have to be made on this behalf, that
is, the distinction between classificatory and relational concepts, and the
distinction between open series and paradigms. Classificatory concepts
– for instance rose and violet – are never in opposition because the open
series they belong to do not contain independent formal relations, where-
as in the realm of relational concepts, open to formal relations, a further
distinction has to be made between purely differential paradigms – for
instance walk, run, hop, skip, crawl – and true oppositions: for instance
good, bad, or hot, warm, lukewarm, cool, cold, icy. The distinction
between exclusive and graded opposition is not so relevant for contra-
diction, whereas relational opposition – for instance servant, master – is
strategic, because it puts classificatory concepts into oppositions.
The combination of all these formal and conceptual parameters
gives birth to a wide typology of expressions spanning from contra-
diction to absence of conflict passing through shallow conflict, from
consistent thinking to the cramp of thought passing through metaphor.
Let us briefly recall the most interesting cases.
The double classification of the same entity does not end in contra-
diction but in shallow conceptual conflict, which is open to metaphor
150 MICHELE PRANDI
in the presence of empirical concepts – This surgeon is a butcher – and

blocked in a cramp of thought when analytical concepts are involved:
This square is a triangle; This figure is a square and a triangle. In the
presence of an opposition between a term and its negation, double classi-
fication typically grades the commitment of the speaker to the classifica-
tion involved: This cottage is not a cottage; This is and is not a cottage.
In the presence of relational opposition, double classification ends in
consistent contradiction, open to the expression of complexity, conflict
and change: John is servant and master; This servant is a master.
The double classification of a relation within the borders of a dif-
ferential lexical paradigm ends in metaphor: Ann’s speaking is singing.
In the presence of opposition, both exclusive and graded, the outcome
is consistent contradiction: This is and is not a battle, This battle is not
a battle; This is and is not joy; Ann’s joy is sadness; Ann’s joy is no joy;
Ann’s joy is sad
When two opposite and consistent relational predicates are jointly
attributed to the same individual subject, the outcome is a true syn-
thetic predication, which is the elective form of contradiction depicting
complex, conflicting and changing situations: Odi et amo (Catulle);
Ardo et son di ghiaccio (Petrarch); I both wished and feared to see
Mr Rochester (Charlotte Brontë); Vorrei e non vorrei (Da Ponte), She
was both frightened and made happy thereby (Tolstoy). In the pres-
ence of differential paradigms, the same form gives birth to a non-
contradictory synthetic predication that plainly describes a composite
object or state of affairs: Ann’s skirt is yellow and red; Ann is speaking
and singing.
After looking deep into this multi-faceted constellation of conflicts,
we are ready to have an overview of the complex relationship between
different forms of conceptual conflict – namely, contradiction, incon-
sistency and shallow conflict – and different figures, that is, oxymoron
and metaphor.
Contradiction never challenges the conceptual identity of the sub-
ject, but simply makes it difficult to grasp the empirical content of a
consistent predicate. When it takes the predicative form as the negation
of a tautology – This cottage is not a cottage; This battle is not a battle
– it challenges the empirical identity of the subject.
An inconsistent metaphor – The moon smiles – challenges the

conceptual identity of the subject, whereas a consistent metaphor –
This surgeon is a butcher – challenges its empirical identity. Owing
to this, both inconsistent and consistent metaphor encapsulate in their
foundation a contradiction of the predicative form – x is not x – which
negates a tautology: the moon that smiles is no longer an inanimate
being; a surgeon that is a butcher is not really a surgeon. Whereas
contradiction of the predicative form directly negates the identity of
a subject, a metaphor presupposes it as a preliminary step in order to
impose a positive and alternative model on the subject: the smiling
moon is no longer the moon because it is seen as a human being; a
given surgeon is not a surgeon because he is a seen as a butcher. Thus,
the contradiction of the predicative form, which negates a tautology,
can be considered as the pars destruens of a conceptual strategy that
has in metaphorical thinking, both inconsistent and consistent, its
pars construens.
Bibliography.
Amari Parker 2006 Dante’s Divine Comedy, Engl. transl. by A. Amari
Parker, London, Arcturus, 2006.
Armstrong 1966 Plotinus, Enneads, I. 8. 1, Engl. transl. by A. H.
Armstrong, in Plotinus, Vol. I: Porphyry on the Life of
Plotinus and the Order of his Books; Enneads, I. 1-9,
:LOOLDP+HLQHPDQQ/7'/RQGRQí+DUYDUG8QLYHUVLW\
Press, Cambridge/Mass. 1966.
Bally 1932 Ch. Bally, Linguistique générale et linguistique françai-
se, Leroux, Paris ; 2nd edition, Francke, Bern, 1944.
Benveniste 1956 E. Benveniste, La nature des pronoms, in For Roman
Jakobson, Mouton, The Hague; repr. in E. Benveniste,
Problèmes de linguistique générale, Gallimard, Paris,
1971.
Bhat 1994 D. N. S. Bhat, The adjectival Category, Amsterdam –
Philadelphia, John Benjamins.
Black 1952 M. Black, Definition, presupposition and assertion,
in «The Philosophical Review», 61, repr. in M. Black,
Problems of Analysis. Philosophical Essays, London,
Routledge & Kegan Paul, 1954, pp. 24-45.
152 MICHELE PRANDI
Bolinger 1967 D. Bolinger, Adjectives in English: attribution and pre-

dication, in «Lingua», 18, pp. 1-34.
Bottiroli 2006 G. Bottiroli, Che cos’è la teoria della letteratura,
Torino, Einaudi.
Burckhardt 1942 J. Burckhardt, Historiche Fragmente. Aus dem Nachlass
Gesammelt von Emel Duerr, Neudruck mit einem
Vorwort von Werner Kaegi, Anstalt – Stuttgart – Berlin,
Deutsche Verlag.
Campbell 1982 Greek Lyric, I, Engl. transl. by D. A. Campbell, The
Loeb Classical Library, London, Harvard University
Press, Cambridge/Mass. – William Heinemann Ltd.
Campbell 1988 Greek Lyric, II, Engl. tr. by D. A. Campbell, The Loeb
Classical Library, London, Harvard University Press,
Cambridge/Mass. – William Heinemann Ltd.
Cardinal 1974 R. Cardinal, Surrealist beauty, in «Forum for Modern
Language Studies», X, 4, pp. 348-356.
Carnap 1932 R. Carnap, Überwindung der Metaphysik durch logische
Analyse der Sprache, Erkenntnis II, pp. 219-241. Engl.
transl. The elimination of metaphysics through logical
analysis of language, in Logical Positivism, A. J. Ayer
ed., New York, Macmillan, 1959, pp. 60-81.
Cellier 1965 L. Cellier, D’une rhétorique profonde: Baudelaire et
l’oxymoron, «Cahiers internationaux de symbolisme»,
8, pp. 3-14.
Chomsky 1965 N. A. Chomsky, Aspects of the Theory of Syntax,
Cambridge / Mass., The M.I.T. Press.
Conte 1988 M. E. Conte, Condizioni di coerenza. Ricerche di lin-
guistica testuale, La Nuova Italia, Firenze. New edi-
tion by B. Mortara Garavelli, Alessandria, Edizioni
dell’Orso, 1999.
Cook 1962 Aristotle, The Categories, Engl. transl. by H. P. Cook in
The Categories; On Interpretation, London – Cambridge/
Mass., Loeb Classical Library.
Cruse 1986 D. A. Cruse, Lexical Semantics. Cambridge, Cambridge
University Press.
Daladier 1978 A. Daladier, Problèmes d’analyse d’un type de nomi-
nalisation en français et de certains groupes nominaux
complexes, Thèse de 3me cycle, Paris, Université Paris
VII, L.A.D.L.
D’Andrade 1987 R. D’Andrade, A folk model of the mind, in Cultural

Models in Language and Thought, D. Holland, N. Quinn
eds., Cambridge, Cambridge University Press, pp. 112-
148.
Dennett 1969 D. Dennett, Content and Consciousness, London – New
York, Routledge & Kegan Paul.
Diderot - d’Alembert D. Diderot, J.-B. d’Alembert, Encyclopédie. Dictionnaire
1772 raisonné des sciences et des arts, Paris.
Dik 1989 S. C. Dik, S. C., The Theory of Functional Grammar.
Part I: The Structure of the Clause, Dordrecht –
Providence. 2nd rev. ed., Mouton De Gruyter, Berlin –
New York, 1997.
Dixon 1977 R. M. W. Dixon, Where have all adjectives gone, in
«Studies in Language», I, 1, pp.19-80.
Dixon 1994 Adjectives, in The Encyclopedia of Language and
Linguistics, R. E. Asher ed., Oxford – New York,
Pergamon Press, pp. 29-35.
Ducrot 1980 O. Ducrot, Les mots du discours, Paris, Les Editions de
Minuit.
Fillmore 1977 Ch. J. Fillmore, Ch. J., Topics in lexical semantics,
in Current issues in Linguistic Theory, P. Cole, (ed.),
Bloomington, Indiana University Press, pp. 76-138.
Gallop 1984 Parmenides of Elea, Fragments, Engl. transl. by D.
Gallop, 7RURQWR í %XIIDOR ± /RQGRQ 8QLYHUVLW\ RI
Toronto Press.
Geeraerts 1988 D. Geeraerts, On necessary and sufficient conditions, in
«Journal of Semantics», 5, pp. 275-291.
Geeraerts 1991 La grammaire cognitive et l’histoire de la sémantique
lexicale, «Communications», 53, pp. 17-50.
Gigliucci 1990 R. Gigliucci, Oxymoron amoris. Retorica dell’amore
irrazionale nella lirica italiana antica, De Rubeis,
Anzio.
Gigliucci 2004 Contraposti. Petrarchismo e ossimoro d’amore nel
Rinascimento. Per un repertorio, Roma, Bulzoni.
Giry-Schneider 1987 J. Giry-Schneider, Les prédicats nominaux en français.
Les phrases simples à verbe support, Droz, Genève.
Grandi 2002 N. Grandi, Morfologie in contatto, Milano, Franco
Angeli.
154 MICHELE PRANDI
Grice 1975 H. P. Grice, Logic and conversation, in Syntax and Seman-

tics. 3: Speech Acts, P. Cole, J. L. Morgan eds., New York
– San Francisco – London, Academic Press, pp. 41-58.
Gross 1987 G. Gross, G. Les constructions converses du français,
Genève – Paris, Droz.
Gross 1993 Trois applications de la notion de verbe support, in
«L’information grammaticale», 59, pp. 16-23.
Haiman 1980 J. Haiman, Dictionaries and encyclopaedias, in
«Lingua», 50, pp. 329-357.
Hamm 1989 A. Hamm, Remarques sur le fonctionnement de la néga-
tion dans les proverbes, in «Recherches Germaniques»,
2, pp. 177-193.
Haser 2005 V. Haser, Metaphor, Metonymy, and Experientialist
Philosophy, Berlin – New York, Mouton De Gruyter.
Hegel 1812-1813 G. W. F. Hegel, Wissenschaft der Logik. Erster Band:
Die Objective Logik, Nürnberg, repr. in G. W. F. Hegel,
Gesammelte Werke, Band XI, Hamburg, Felix Meiner
Verlag, 1978. Engl. transl. Hegel’s Science of Logic,
George Allen & Unwin Ltd., New York, London –
Humanities Press,.
Hjelmslev 1943 L. Hjelmslev, Omkring sprogteoriens grundlaeggelse,
Copenhagen. Engl. transl.: Prolegomena to a Theory of
Language, Madison, The University of Wisconsin Press,
1961.
Humboldt 1836 W. von Humboldt, Über die Verschiedenheit des men-
schlichen Sprachbaues und ihren Einfluss über die
Geistliche Entwicklung des Menschengeschlechts,
Berlin. Repr. in W. von Humboldt, Werke in fünf Bänden,
Band III, Darmstadt, Cotta, 1963.
Husserl 1901 E. Husserl, Logische Untersuchungen. Band 1:
Halle 1900; Band II: Halle 1901. Rev. Ed.: Band I
(Prolegomena), Band II, Teil I (Res. I - V), Halle 1913;
Band II, Teil II (Res. VI), Halle 1921. Critical ed.:
Husserliana, Vol. XVIII (1975) - XIX, I - II (1984),
Nijoff, The Hague. Engl. transl.: Logical Investigations,
Routledge & Kegan Paul, London, 1970.
Innes 1995 Aristotle, Poetics, Longinus, On the Sublime, Demetrius,
On Style, On the Sublime, X, Engl. transl. by D. C. Innes,
Harvard University Press, The Loeb Classical Library,
Cambridge/Mass. – London.
Jones 1959 Heracleitus, On the Universe, Engl. transl. by W. H.

S Jones in Hyppocrates, Vol. IV, Heracleitus, On the
Universe:LOOLDP+HLQHPDQQ/7'/RQGRQí+DUYDUG
University Press, Cambridge/Mass.
Kant 1763 I. Kant, I., Versuch den Begriff der negativen Größen
in die Weltweisheit zu führen, Königsberg. Repr. in
Immanuel Kants Werke, Band II, Berlin 1912. Engl.
transl., Attempt to introduce the concept of negative
magnitudes into philosophy, The Cambridge Edition
of the Works of Immauel Kant, Theoretical Philosophy,
1755-1770, Cambridge, Cambridge University Press,
1992.
Kant 1781 Kritik der reinen Vernunft, I. Riga, 1781. II. Riga, 1787.
Repr. in: Immanuel Kants Werke, Band III, Berlin, 1913.
Engl. transl., Critique of Pure Reason, Macmillan,
London, 1963.
Kleiber 1990 G. Kleiber, La sémantique du prototype, Paris, Presses
Universitaires de France.
Kleiber 1998 Les proverbes antinomiques: une grosse pierre ‘logi-
que’ dans le jardin toujours ‘universel’ des proverbes,
in Le figement lexical, S. Mejri, G. Gross, A. Clas, Taïeb
Baccouche (eds.), Tunis, pp. 51-75.
Lakoff 1971 G. Lakoff, Presuppositions and relative well formed-
ness, in Semantics, D. D. Steinberg, L. A. Jakobovits
eds., Cambridge, Cambridge University Press, pp. 329-
340.
Landheer 1996 R. Landheer, Le paradoxe: un mécanisme de bascule,
in Le paradoxe en linguistique et en littérature, R.
Landheer, P. J. Smith eds., Genève, Droz, pp. 91-116.
Leishman 1954 Selected Poems of Friedrich Hölderlin, Engl. transl. by
J. B. Leishman, London, The Hogarth Press.
Lepschy 1981 G. Lepschy, Freud, Abel e gli opposti, in G. Lepschy,
Mutamenti di prospettiva nella linguistica, Il Mulino,
Bologna.
Lepschy 1989 Enantiosemia e ironia nel lessico italiano, in G. Lepschy,
Nuovi saggi di linguistica italiana, Bologna, Il Mulino.
Lounsbury 1964 F. L. Lounsbury, The structural analysis of kinship
semantics, in Proceedings of the Ninth International
Congress of Linguists, H. G. Lunt, The Hague, Mouton,
pp. 1073-1093.
156 MICHELE PRANDI
Lyons 1963 J. Lyons, Structural Semantics, Oxford, Blackwell.

Lyons 1977 Semantics, Cambridge, Cambridge University Press.
McCawley 1970 J. D. McCawley, Where do noun phrases come from?,
in Readings in English Transformational Grammar,
R. Jakobs, P. S. Rosenbaum eds., Waltham / Mass.,
Blaisdell. Rev. version in Semantics, D. D. Steinberg, L.
A. Jakobovits eds., Cambridge, Cambridge University
Press, pp. 217-231.
Molinié 1992 G. Molinié, Dictionnaire de rhétorique, Paris, Livres
de Poche.
Morier 1961 H. Morier, Dictionnaire de poétique et de rhétorique,
Paris, Presses Universitaires de France.
Mortara Garavelli 1985 B. Mortara Garavelli, La parola d’altri. Prospettive di
analisi del discorso riportato, Palermo, Sellerio.
Most 2006 Works and Days, Engl. transl. by G. W. Most in Hesiod,
Theogony, Works and Days, Testimonia, The Loeb Classical
Library, Cambridge/Mass., Harvard University Press.
Murray 1995 Homer, The Odyssey, 19, Engl. transl. by A. T. Murray,
rev. by G. E. Dimock, The Loeb Classical Library,
Cambridge/Mass. – London, Harvard University Press.
Murray 1999 Homer, Iliad, Engl. transl. by A. T. Murray, rev. by W.
F. Wyatt, The Loeb Classical Library, Harvard/Mass. –
London, Harvard University Press.
Musa 1996 F. Petrarca, The Canzoniere, or Rerum vulgarium frag-
menta, Engl. transl. by M. Musa, Bloomington/Ind.,
Indiana University Press.
Nida 1975 E. A. Nida, Principles of translation as exemplified
by Bible translating, in On Translation, R. A. Brower
ed., Cambridge / Mass., Harvard University Press, pp.
11-31.
Ossola 1977 C. Ossola, Apoteosi e ossimoro, in «Rivista di storia e
letteratura religiosa», XIII, pp. 47-103.
Palmer 1976 F. R. Palmer, Semantics, Cambridge, Cambridge
University Press.
Perrin 1996 L. Perrin, L’ironie mise en trope, Paris, Editions Kimé.
Porzig 1943 W. Porzig, Wesenhafte Bedeutungsbeziehungen, in
«Beiträge zur deutschen Sprache und Literatur», 58, pp.
70-97.
Poublanc 1991 F. Poublanc, Logique et métaphysique, in «Revue de

métaphysique et de morale», 2, pp. 147-179.
Prandi 1987 M. Prandi, Sémantique du contresens, Paris, Les Editions
de Minuit.
Prandi 1992 Grammaire philosophique des tropes, Paris, Les Editions
de Minuit.
Prandi 2004 The Building Blocks of Meaning, Amsterdam –
Philadelphia, John Benjamins.
Prandi 2007 Conceptual conflict and metaphor: against literal mea-
ning, in Metaphors, Concepts and Cognition, S. Arduini
ed., Roma, Edizioni di Storia e Letteratura.
Rosch 1972 E. Rosch, Universals in Color Naming and Memory,
«Journal of Experimental Psychology», 93, pp. 10-20.
Rosch 1973 On the internal structure of perceptual and semantic
categories, in Cognitive development and the Acquisition
of Language, T. E. Moore ed., New York – San Francisco
– London, Academic Press, pp. 111-144.
Rosch 1975 Cognitive reference points, «Cognitive Psychology», 7,
pp. 532-547.
Rosch 1978 Principles of categorization, in Cognition and
Categorization, E. Rosch, B. B. Lloyd eds., Hillsdale /
N. J., Lawrence Erlbaum Associates, pp. 27-48.
Sapir 1944 E. Sapir, On grading: a study in semantics, in «Philosophy
of Science», 2, pp. 93-116. Repr. in E. Sapir, Selected
Writings in Language, Culture and Personality, ed. by
D. G. Mandelbaum, The University of California Press,
Berkeley, 1949, pp. 122-149.
Saussure 1916 F. de Saussure, Cours de linguistique générale, Paris,
Payot. Critical ed. by T. De Mauro, Paris, Payot, 1972.
Engl. transl. Course in General Linguistics, Fontana /
Collins, London, 1974.
Searle 1983 J. Searle, Intentionality, Cambridge, Cambridge Univer-
sity Press.
Sommers 1963 F. Sommers, Types and ontology, in «The Philosophical
Review», 72, pp. 527-563. Repr. in Philosophical Logic,
P. F. Strawson ed., Oxford, Oxford University Press, pp.
138-169.
Sperber - Wilson 1986 D. Sperber - D. Wilson, Relevance. Communication and
Cognition, Oxford, Blackwell.
158 MICHELE PRANDI
Steinberg - Jakobovits D. D. Steinberg - L. A. Jakobovits, Semantics, Cambridge,

1971 Cambridge University Press.
Strawson 1952 P. F. Strawson, Introduction to Logical Theory, London,
Methuen.
Strawson 1959 Individuals. An essay in Descriptive Metaphysics,
London, Methuen & Co, 2nd ed. 1964.
Taylor 1989 J. R. Taylor, Linguistic Categorization: Prototypes in
Linguistic Theory, Oxford, Clarendon Press.
Trier 1931 J. Trier, Der deutsche Wortschatz im Sinnbezirk des
Verstandes. Die Geschichte eines sprachlichen Feldes.
Part I: Von den Anfangen bis zum Beginn des 13.
Jahrhunderts, Winter, Heidelberg. Repr. in J. Trier,
Aussätze und Vorträge zur Wortfeldtheorie, herausgege-
ben von A. van der Lee, O. Reichmann, Mouton, The
Hague – Paris, 1973, pp. 40-65.
Vendler 1970 Z. Vendler, Say what you think, in Studies in Thought
and Language, J. Cown ed., Tucson, The University of
Arizona Press, pp. 79-97.
Weber 1922 M. Weber, Wirtschaft und Gesellschaft, Mohr,
Tübingen.
West 1993 Greek Lyric Poetry, Engl. transl. by M.L. West, Oxford
World’s Classics, Oxford, Oxford University Press.
Wierzbicka 1980 A. Wierzbicka, Lingua Mentalis, Sidney – New York,
Academic Press.
Wierzbicka 1987 Boys will be boys: ‘Radical Semantics’ vs ‘Radical
Pragmatics, in «Language» 63, pp. 95-114.
Wierzbicka 1996 Semantics. Primes and Universals, Oxford – New York,
Oxford University Press.
Wittgenstein 1929 Some remarks on logical form, in «Proceedings of the
Aristotelian Society», Suppl. Vol. 9, pp. 162-171.
Wittgenstein 1960 Tagebücher 1914-1916, in L. Wittgenstein, Tractatus logi-
co-philosophicus. Tagebücher 1914-1916. Philosophische
Untersuchungen, Frankfurt am Main: Suhrkamp. Engl.
tr. Notebooks 1914-1916, Blackwell, Oxford, 1961.
Wittgenstein 1969 Über Gewissheit – On Certainty, Blackwell, Oxford.
Polus - Rethorica.qxp 09/09/2011 17.53 Pagina 247
POLUS
RETHORICA
1. Metaphors, edited by STEFANO ARDUINI, 2007, pp. 120.
2. Paradoxes, edited by STEFANO ARDUINI, 2011, pp. 160.

Finito di stampare nel settembre 2011
dalla GRAFICA EDITRICE ROMANA srl
Via Carlo Maratta, 2/b - Roma
Tel./Fax 06.57.40.540
graficae1@graficaeditriceromanasrl.191.it

Paradoxes (Labov)

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Paradoxes (Labov)

Uploaded by

Copyright:

Available Formats

principi.qxp 09/09/2011 17.

EDIZIONI DI STORIA E LETTERATURA

Prima edizione: settembre 2011

Tutti i diritti riservati

EDIZIONI DI STORIA E LETTERATURA

Giambattista Vico in his works writes that tropes are rhetorical

In his book Novantiqua Paolo Valesio1 concludes that antithesis

Obviously, the subject of antithesis has many rhetorical sides: nega-

Humans are beings who cannot distinguish between dreaming and

descriptive inadequacy. Only mythical and paradoxical language can

dissolves the contradiction by dissociating personal responsibilities:

culture, of Platonism and other forms of abstract idealities. The soph-

AN INTRODUCTION TO (CONJUNCTIVE) SCISSIONAL LOGIC

1. Can a unified definition of paradoxes be given? Can a common

rendered transparent by any of its interpreters. What are its principles,

2. Initial ambiguity needs to be avoided: conjunctive logic is not

Very important consequences immediately derive from this parado-

In the conjunctive regime, the opposition between rigidity and flex-

3. Many scholars see paradoxes as providing the best confirmation

discovering that, alongside ‘chained’ paradoxes, there are others whose

4. We appear to have taken on a ‘symtomal interpretation’ not only

unacceptable. In opposition to this, one could well provide the follow-

5. What I shall now present is a conceptual research programme,

so widespread as to be difficult to perceive – unless one takes a step

philosophers who find logical inspiration in literature, and, more gen-

6. Let us take a quick glance at the paradoxes discussed especially

mentary typologies originated, such as Ramsey’s division between set-

thoroughly explore the hypothesis that chained paradoxes, i.e. aporias,

7. It is not difficult to find immediate confirmation: for example, in

a single paradox. If we manage to clarify its mechanism better than

I was by no means surprised to learn that in a group of people who were

8. Let us admit, for the moment, that Tarski’s conception is error

denies the autonomy of meaning. This fallacy is opposed by what I

F. Nietzsche, Nachgelassene Fragmente 1885-1887.

Essays on the Theory of Sexuality, 1905.

by Kripke29. Let us consider these two utterances, made by Jones and

no language can become an object language (in the separative sense)

by taking on routine features, resembling the need to devote time to

Why does Watson’s answer sound funny? Not because it is false, or

9. It is worth pausing briefly over the limits of the pragmatic turn

which symbolic logic is unable to analyse, even when dealing with

not the relationship between a master and slave. No logical transgres-

11. We have taken an important step forward in our research, and

ships. The principle of non-contradiction presides over all relationships

Contradictories are by no means the strongest case of opposition,

12. Let us imagine, however, an interlocutor who does not want

Among the relationships envisaged in the logical square, that

the places it provides for. In the case under discussion, as a matter of

non hatred non love

quences of this obstinate inclination to compromise requires some

No, because language is flexible enough to allow twisting and intensi-

Confrontation, exchange of ideas between specialists in various

Aristotle, Categories 7, 7 b 15;

man, for example, that animal?” Aristotle, ibidem, 5, 3b 24-25.

14. Correlatives are the obstacle or rather adversary in the way of

Let us return to the question so far unanswered: is it always possible

16. We use this term in the psychoanalytical sense. In Freud’s view,

The concept of identification upsets the classic theory of predica-

17. At this stage we must ask ourselves what the limits to be

rigid identity is situated, on the other hand, in the sphere of separative

use traditional psychology of faculties). Thus Zerlina’s “Vorrei e non

19. Here follows a list of theses, considering the route followed so

– ancient (Aristotle) and modern (Frege, etc.) logic turned to a single