ELISEO VERON

Ideology and Social Sciences: A Communicational

Approach

1. MEANING AND PRAGMATICS

One of the obvious levels of empirical reality of what we call science

corresponds to the PRODUCTS of scientific activity: bodies of linguistic
materials consisting, in different proportions, of propositions made up
of signs of a natural language and elements of artificial or formal
languages. With regard to any system of signs, there is a traditional
distribution of areas: (a) SYNTACTICS, the study of the relationships
among signs themselves. Syntactics has been characterized as the study
of the rides for constructing "acceptable" expressions within a given
language system, irrespective of their meanings, (b) SEMANTICS, the study
of how signs are related to what they stand for, refer to, or 'represent'.
Semantics is supposed to state the rules of correspondence between
signs and their denotata, (c) PRAGMATICS, the study of the relationships
between signs and their human users, i.e., those who send and receive
the signs available in particular situations. From this point of view,
therefore, the science of sign-phenomena comprises three basic chapters:
the analysis of the formal relationships between signs, the analysis of
meaning, and the analysis of use.1
An empirical science considered as a linguistic system and studied
without taking into account the associated processes relating signs to
the communicators, could be described by the set of its syntactic-seman-
tical construction rules. As, for the time being, the social sciences use
1
See R. Carnap, Introduction to semantics (Cambridge, Harvard University Press,
1946).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
60 ELISEO VERON

natural languages in a much greater proportion than artificial languages,

so, the specialization of natural languages within social sciences requires
the introduction of additional rules, syntactic as well as semantic, aimed
at reducing the high degree of ambiguity of the 'non-scientific' use of
natural languages.
Pragmatics begins when this whole set of rules is considered as a body
of norms of procedure that a given user of the system applies under certain
empirical conditions of communication. In any particular situation, the
emission (construction) and reception (consumption) of a body of signs
require the application of some of these rules. What is usually called
'methodology of social research' is the normative body of syntactic-
semantical rules that any user of the system has to take into account
when engaged in the scientific process leading to a final result: the emis-
sion of signs consisting of a (supposedly) adequate description, meaning-
ful to other users of the system, of certain properties of a given set of
empirical facts. A report, an article, a book: these are the objective
forms, the empirical reality of what we call 'knowledge'. Under a most
important perspective, science exists empirically as an open CORPUS
OF TEXTS.
Given a set of syntactic-semantical rides, any emission or reception
operation within a given context of communication is a decision-making
process. This means that although the syntactic-semantical rules define
a certain field of restrictions for the users of the system, there are always
several degrees of freedom or indetermination. The body of syntactic-
semantical rules is never a deterministic or closed set of rules. As every-
body knows, the decisions underlying the production of any set of signs
aimed at a 'scientific' description of reality imply a complex system of
operations: selection of concepts; formulation of hypotheses; opera-
tionalization; model construction; building of instruments for collecting
evidence; data analysis according to rules with different degrees of
standardization; and so on. The study of the empirical conditions under
which these decisions take place corresponds to the pragmatics of science.
In social sciences, there is a typical repertoire of problems that most
of the time are discussed as if pragmatics did not exist: scientific objec-
tivity; the role of value judgments; the relationships between science
and ideology and other related issues. These problems have been almost
exclusively approached from a syntactic (logical) and/or semantic
(epistemological and methodological) point of view. It is the central
contention of this paper that these andother crucial problems can be
satisfactorily formulated only at the level of pragmatics.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 61

The fact that most speculations around scientific objectivity in the

social sciences have excluded pragmatical considerations stems from
the particular theory of meaning associated with the logical background
of positivism. In fact, the classical triad of syntactics, semantics, and
pragmatics is built entirely upon a basic assumption: the description
of the formal rules of a language system, plus the study of the rules of
correspondence between signs and things, give us a complete picture
of the process of meaning; the remaining area, that of the 'use' of the
system, involves empirical problems that are 'external' to the basic
syntactic-semantical model. If this may perhaps be true in reference
to artificial languages, it is obviously wrong when a natural language
(and therefore processes of social communication) is involved. This
does not mean that the syntactic-semantical levels of analysis are useless
when dealing with social systems of communication; it simply means
that syntactic-semantical descriptions are partial or incomplete descrip-
tions of meaning processes. From the point of view of human communica-
tion theory, meaning is a social process. Syntactics and semantics, as
they have been classically defined, state SOME of the general conditions
to which this process is subject. Pragmatics is not the study of how signs
(and therefore meanings) are 'used'; it describes how meaning comes
into being as an empirical fact of social communication.
Since a fully developed theory of social communication is not available,
the tendency has been to try to solve pragmatic problems (like those of
scientific objectivity, the alleged freedom from value judgments, and the
relationships between science and ideology) within the narrower limits
of syntactic-semantical considerations. This situation has led, in fact,
to an impossible task, because the pragmatic issues involved in scientific
practice cannot be decided in terms of syntactic-semantical rules of
procedure.
Let me express this idea in terms of a distinction which, under one
form or another, has been very important in the development of contem-
porary linguistics, and has been recently reshaped by Chomsky as the
difference between COMPETENCE anp PERFORMANCE. Competence refers
to the knowledge we must impute to the 'normal' speaker of a language
in order to explain his capability to produce an infinite number of
"correct" utterances of that language. "A grammar, in the traditional
view, is an account of competence."2 Performance, on the other side,
is what the speaker DOES when he uses his native language system.
2
N, Chomsky, Topics in the Theory of Generative Grammar (The Hague, Mouton
& Co, 1966), 10.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
62 ELISEO VERON

Both aspects are of course related to one another, at least by virtue of

the fact that "performance provides evidence for the investigation of
competence".3 But the relationship between competence and performance
is not entirely clear, even within the limits of the transformational
models of generative grammar. One of the current assumptions under-
lying many discussions about scientific objectivity in the social sciences
has been to suppose that the decision-making process that is at work
at the level of actual scientific performance can be directly inferred
from the existing model of scientific competence. But in fact there exists
a gap between the two. In order to build a bridge leading from scientific
competence to scientific performance we need a theory of human com-
munication. This implies a rejection of the definition of pragmatics as
the study of the relationships between signs and their users. I would
propose that the theory of pragmatics has to cover the complex and
unexplored field that begins BEYOND DENOTATION. We do not have such
a theory yet. But let me try to show that we need it, in this particular
case, in order to reach a better understanding of the boundaries between
science and ideology. The recent developments of semiotics, under the
impact of contemporary linguistics and a broader approach to com-
munication processes, imply a new look at the problems of scientific
language.

2. METACOMMUNICATION

In what follows, we shall be concerned with linguistic discourse produced

and consumed with a predominantly REFERENTIAL function, i.e., linguistic
material that is supposed to describe in a certain way the extralinguistic
universe of reality. Scientific discourse belongs to a subtype of this class
of messages.
Let us take a sort of science-fiction illustration. Suppose a world in-
habited by intelligent beings, with a communication system such that
the pragmatic analysis of their communications would be uninteresting:
a syntactic-semantic analysis would explain all meaningful properties
of the messages transmitted. In other words: the model of competence
would explain in an exhaustive way the phenomena of performance.

3
Ibid. See also N. Chomsky, Aspects of the Theory of Syntax (Cambridge, Mass.,
The M.I.T. Press, 1965) and Current Issues in Linguistic Theory (The Hague, Mouton
& Co., 1967).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 63

Communication in such a world would have to fulfill the following

conditions:
(a) The language used by those beings would be strictly formal; we
can imagine an entirely formalized digital system wherein the relation-
ships between signs and denotata are strictly arbitrary.
(b) Communication would have to take place through a single
channel; i.e., there exists only ONE informational series between com-
municators.4
(c) There would have to be no possible variations of the signals. If these
beings communicate through visual signals, there has to be one and only
one way of representation of each signal, without variations in size,
shape, etc.; if we imagine an auditory system, there have to be no
variations of tone, volume, pitch, etc., that could affect referential
meaning.
(d) There would have to be no body language of any kind.
(e) The syntactic rules of composition of messages would have to be
strictly deterministic at all levels of complexity; the user has no options
concerning the combination of signals to build a message.
(f) The semantic rules would have to be strictly biunivocal; i.e., for
each thing, quality, or process that may be referred to, there would be
one and only one sign or combination of signs to denote it, and each
sign or combination of signs would correspond to one and only one
denotatum in reality. So, for each possible message content, the sender
would have one way and only one to transmit it, and the receiver would
have one and only one way to interpret.
(g) The reciprocal perception of the communicators would have to
be void of temporal dimension (i.e., it would have to be 'instantaneous')
and each receiver would have to be able to codify the content
of the messages but not the identity of the sender. Were it not so,
the sole presence of one communicator before another, even if there
were no transmission of information through the existing channel,
would convey non-denotative meanings.
(h) Contextual rules, i.e.9 the rules concerning what is communicated
in what situation, would also have to be deterministic.
(i) The rules of relationships between communications and silences
would also have to be deterministic; that is to say, silence would have
no meaning.
4
For a typology of series of informational events, see E. Veron, "Comunicacion
y trastornos mentales: el aprendizaje de estructuras", Acta Psiquiatrica y Psicologica
de America Latina, 10: 77-85 (1964).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
64 ELISEO VERON

(j) In each communicational situation between two beings of this

world,5 the content of the messages would be unknown to the receiver
before communication takes place.
Condition (j) is necessary to avoid the consequence of this world being
entirely redundant, in which case communication acts would be useless.
Therefore, we do not need to identify our intelligent beings as gods;
they must have some degree of ignorance or, in other terms, we suppose
that, in spite of their intelligence, they live in an empirical reality that
still has sectors unknown to them.
Let us consider this imaginary example, regardless of whether these
conditions make sense empirically, i.e., whether they are — in the weber-
ian sense — Objectively possible'. The important point is that in such a
world the science of communication would have only two interesting
chapters: syntactics and semantics. For all practical purposes, the
messages would have only a denotative function; connotation would
not exist or would be reduced to a minimum. In human communication,
of course, this does not happen: the meaning of a message goes far beyond
its denotative function. All human messages denote at one level, and
connote at other levels. Thanks to this, no doubt, we live in a world
much less boring than the one we have described or — what amounts
to the same — we live in a much more complex and ambiguous com-
munication universe.
Connotation is a level of meaning that exists as far as the rules of
message construction are not entirely deterministic in the different
dimensions of human communication. Therefore, the sender has, in
each concrete situation, a number of ALTERNATIVES open to him for
constructing the messages, and these options are not decidable in terms
of the syntactic-semantical rules of the system. A sender within a system
of communication with certain degrees of freedom performs two basic
operations to send a message: among the repertoire of units composing
the code of the system he SELECTS those that will compound the message,
and he COMBINES the selected units in a certain way within the message.6
The connotative meaning of the message, i.e., its metacommunicational
dimension, depends then on the selective and combinatory options at
the disposal of communicators.
Prieto has referred to the manner according to which a given operation
5
In terms of the conditions stated, it would amount to the same whether we imagine
that this universe is inhabited by two communicators or any number of 'people'.
6
R. Jakobson, in Part Two of R. Jakobson and M. Halle, Fundamentals of Lan-
guage (The Hague, Mouton & Co., 1956).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 65

is executed, "as far as it is not the only possible manner", and he has
called it STYLE.7 Prieto's conception of 'style9 coincides with what we
call here, from the point of view of information transmission, meta-
communication. The manner according to which an operation (denota-
tion) is carried on, manifests itself through the selective and combinatory
decisions made by the sender. It makes sense to speak of a decision-
making process only when there is MORE THAN ONE POSSIBLE WAY of
denoting some set of extralinguistic entities. This decision-making
process generates messages having TWO LEVELS OF MEANING. On the one
hand, we have the message composed by linguistic signs fulfilling a
denotative function. Each of them may be analyzed into significans and
significatum. This is the level of COMMUNICATION. On the other hand,
the fact that the sender has made these particular decisions concerning
selection and combination, within this particular communicational
situation, creates a second level of meaning: the field of connotation.
This field comprises what may be called the METACOMMUNICATIONAL
phenomena. The notion of metacommunication refers to a process of
transmission of information concerned not with the content of the
messages, BUT WITH THE SELECTIVE AND COMBINATORY OPERATIONS
MADE BY THE SENDER.8
Ashby offers a classical example directly related to the fundamental
principle of metacommunication.9 Two countries, A and B, are at war.
Each of them takes a soldier of the other as a prisoner. Some time after,
the wives of both prisoners receive the following message:
"I am well"
It is known that in country A the prisoners may send to their families
one of the following messages:
"I am well"
"I am slightly ill"
"I am seriously ill"
7
Luis J. Prieto, "Lengua y connotation", in E. Veron (ed.), Lenguaje y comunicacion
social (Buenos Aires, Nueva Vision, 1969). See also Messages et signaux (Paris,
PUF, 1966), 164-65.
8
In many papers, Gregory Bateson and his co-workers in the USA have developed
the concept of metacommunication. See in particular Bateson's chapters in J. Ruesch
and Gregory Bateson, Communication: The Social Matrix of Psychiatry (New York,
Norton, 1951), and P. Watzlawick, J. Beavin, and D. Jackson, Pragmatics of Human
Communication (New York, Norton, 1967). An analysis of connotative processes as
they relate to ideology has been carried out by Roland Barthes. See especially his
paper "Le my the, aujourd'hui" in: Mythologies (Paris, Editions du Seuil, 1957).
9
W. Ross Ashby, An Introduction to Cybernetics (London, Chapman & Hall, 1956).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
66 ELISEO VERON

In country B, only one message is authorized to the prisoners:

"I am well"
which simply means "I am alive", the other alternative being the absence
of any message.
It is evident that the messages received by the soldiers' wives, though
identical in their communicated content, are very different in their
metacommunicational value. WE CAN THEREFORE SAY, FOLLOWING ASHBY,
THAT 'MEANING' is NOT AN INTRINSIC PROPERTY OF A MESSAGE, BUT
DEPENDS ON THE SET OF ALTERNATIVES THE MESSAGE COMES FROM. Any
message determines its connotative meaning in a given situation in relation
to other messages that could have been transmitted instead, and in relation
to different combinations of the same elements integrating the message.
All this may seem so general as to be useless. Notwithstanding, it
must be taken into account that metacommunication is a pragmatic
notion, i.e., it refers to a process of meaning that can only be studied in
relation to the empirical communication system: sender, receiver, and
the concrete situation in which the transaction takes place. Otherwise,
the repertoire of possible alternatives for a given verbal message would
be the whole set of messages that may be constructed within the system
of a given language. It is well known that this set is, for all practical
purposes, infinite, and so any intent of establishing the metacommuni-
cated meanings would be a hopeless task: instead of the emitted message,
we could have simply found any other imaginable message in the same
language. Pragmatical considerations, which include the identity of the
communicators, their social characteristics as well as the social nature
of the communicational relationship existing between them, and also
the subject matter the communication is about, if we are studying messages
centered in the referential function, drastically reduce the potential
universe of discourse for a given situation.
Let me introduce some additional examples corresponding to different
levels of social communication, in order to clarify the functioning and
significance of metacommunicational processes. The establishment of
new interpersonal relationships (formal or informal) is a typical case
where connotation acquires a particular value. The exchanged messages
transmit something through their content, but at the same time they
orient each receiver in the understanding of how the other communicator
interprets the interpersonal relationship. If I greet a person I have just
been introduced to and in doing so I use a tone of voice and terms
implying a certain familiarity, I am metacommunicating, besides trans-

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 67

mitting a greeting, the decision I have taken concerning the future

character of the relationship; this metamessage is a sort of proposal: "I
am ready to engage in a cordial, relatively close relation with you". This
message exists by virtue of the fact that I could have chosen other words
and I could have combined them in a cool and formal fashion, but instead
I have made a different decision. In these cases, each message through
metacommunication proposes NORMS for the future development of the
relationship. This function is carried out as far as each participant
metacommunicates that he has decided to send THAT message and to
exclude several others he could have selected, and therefore that the
message transmitted is for him the 'adequate' one.10
In cases of ongoing interpersonal relationships, the mechanism of
metacommunication becomes evident in those circumstances when the
already established norms of transaction are suddenly altered. If a friend
whom I am seeing every day greets me one morning using an unusual
phrase, this will provoke my surprise and perhaps my confusion. Why,
instead of his usual greeting, has he said this or that, or talked that
way, or kept silent? This question is not directed toward the message
content, but toward coding operations.
The case of synonymy is one of the simplest examples mentioned by
linguists in connection with the selective operations of the sender. In
Argentina there exists a humorous mass-media personage, Mirna
Delma.11 Where her cousins speak of thieves, Mirna Delma says "pol-
troons"; where everybody would speak about horse races, Mirna Delma
refers to a "cavalry competition". Both kinds of discourse are identical
with regard to the communicated content but the personage and the
facetious effect are based on the difference established within a certain
synonymic field stemming from the selective operations. Mirna Delma's
speech metacommunicates her candid search for prestige, her snobbery,
her vulgarity, and, in the last instance, her marginality with regard to
the in-group culture of her social class.

3. SCIENCE AND IDEOLOGY

At the societal level, connotation is the communicational dimension

through which ideological metamessages are transmitted embodied in
10
This point has been developed within the context of a communicational theory
of neurotic disorders in: E. Veron and Carlos E. Sluzki, Comunicacion y neurosis
(Buenos Aires, Editorial del Institute, 1969).
II
Created by Landru, an Argentine humorist.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
68 ELISEO VERON

messages having a manifest referential function. Just as in interpersonal

relationships communicators transmit through metacommunication the
'image' they have with regard to the ongoing relationship and its norms,
so the social mass-messages always metacommunicate a certain 'image'
of society, a certain conception of social reality, the way of organizing it,
and the way of understanding its different aspects. As this image and
these ways of conceiving and understanding social reality are not the
only possible ones, and as they are transmitted through metacommunica-
tion, /.e., at an implicit level of meaning, the term IDEOLOGICAL COM-
MUNICATION seems fairly adequate.12
Let us take a computer analogy. Suppose a machine able to produce
ideological material of a given type: we must NOT identify ideology with
the Output' of the machine. This means that an ideological system is
NOT a body of propositions of a certain kind, but instead a set of semantic
rules defining the constrictions to which the production of a certain
kind of propositions is subject. The set of propositions that may be
generated by the rules is, for all practical purposes, infinite. The set of
rules is a finite set representing the semantic properties of a given semantic
universe: they state the decisions about selection and combination of the
semantic units needed to construct the messages. Therefore, our ideo-
logical system is not the Output' of the computer but its PROGRAM.
From this point of view, then, and at this level of analysis, an 'ideology'
may be defined as A SYSTEM OF SEMANTIC RULES TO GENERATE
MESSAGES.13
The consequences stemming from the operation of different 'semantic
programs' in the field of ideology can be clearly seen when studying how
the relevant social facts are incorporated into mass-media messages, a
process we have called elsewhere SEMANTIZATION.U Suppose an event
taking place within society at a given moment. We shall call it 'X'.
In this particular case (and my own description is itself a process of
semantization) it was the assassination of a Union leader in Buenos Aires
in 1966. Two mass-media of the 'newsweekly' type included information
about it. These were the headings:

12
If we define ideology as a level of meaning of messages manifestly centered in
the referential function, PROPAGANDA may be distinguished from ideology as the
discourse centered in the conative function. On the functions of linguistic messages,
see Jakobson, Essais de linguistique generate (Paris, fiditions de Minuit), chapter 11.
13
See E. Veron, "Ideologia y comunicacion de masas: la semantizacion de la
violencia politica", in Lenguaje y comunicacion social.
14
Ibid.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 69

Mass-medium A Mass-medium B
(middle-class circulation) (working-class circulation)
The Nation:
The first gunfires Political assassination

The middle-class magazine has internal permanent sections, and the

note about the assassination was the first of one of them: 'The Nation'.
The other magazine has no internal permanent structure. If we look at
these headings in some detail, we may find several interesting facts.
Mass-medium A: The combination of the section title with the heading
carries a metamessage that we may express tentatively in the following
way: 'You will read something that has national significance, something
that has happened this week within the context of the Nation'. (In fact
there is in this magazine an implicit but easily detectable rule consisting
in putting the most important event of the week in the first place within
The Nation' section). Let us call this operation CONTEXTUALIZATION.
Second, 'first' as a semantic unit belongs to a temporal paradigm:
something will follow. The fact you will read about is the first mani-
festation of something that is going on, and that will have some other
manifestations later. In a word: the fact you will read about is not an
isolated one; it belongs to a process. Let us call this operation TEM-
PORALIZATION. It must be remarked that if we talk about 'the first gun-
fires', we are referring to a conflict that began before, and is now assuming
a more violent character. And by virtue of the fact that if we talk about
the FIRST gunfires we are supposing that there will be more of them,
we can also argue that there is something like a PREDICTION involved.
Mass-medium B: What about the other magazine? The operation under-
lying the heading is transparent: it says that there are in society political
events; that there are also criminal events, and that we are facing an
intersection of these two classes: this intersection defines the event we
will read about. Let us say that this operation is a sort of CLASSIFICATION.
Within the narrow limits of two- or three-word headings, the two
mass-media have begun to give structure to two semantic universes
that we may suspect will be very different:15 one of them has CONTEX-
TUALIZED and TEMPORALIZED the fact 'X' and has made a PREDICTION;
the other has simply CLASSIFIED the fact 'X' by means of a class inter-
section, an intersection of semantic fields. These operations are the kind
15
A more detailed analysis of connotative mechanisms in mass-media, from which
this example has been extracted, may be found in E. Veron, "Ideologia y comuni-
cacion de masas...", he. cit.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
70 ELISEO VERON

of mechanisms involved in the decision-making process taking place at

any source in social communication, when we look at the construction
rides underlying their messages.
An important consequence of what has been said is that the difference
between 'science' and 'ideology' has not to be understood as a difference
between two types of language or discourse. The difference is one of
LEVELS OF MEANING of the messages circulating in society. Since selection
and combination (which create the connotative dimension of language)
are PRAGMATIC concepts referred to the decision-making process operating
in the source, that conclusion radically modifies the basis of many
traditional discussions around the problem of objectivity.
In science, the ideological level of meaning stems from all those options
in the construction of scientific language that are not decidable in terms
of the formal rides of scientific procedure. This field, as everyone knows,
is very wide in the social sciences today. When carrying out investigations,
social researchers must choose between stratification models where
boundaries between social groups are conceptualized as quantitative,
and social class models; between a concept of mental illness as a deviation
from the existing social norms and a concept of mental disturbance as
a form of adaptation to social contradictions; between a theory of social
development cast in the traditional/modern continuum, and a theory
excluding the idea of a linear progression towards increasing 'rationality';
between a functional model of total society focused on consensual
mechanisms and the reciprocal adjustment of social institutions, and a
model emphasizing conflict areas, and so on. At the present state of the
social sciences, the theoretical and conceptual decisions concerning these
alternatives (and many others) make room for 'sociological discourses'
having an ideological dimension: there exist neither logical-methodolo-
gical nor empirical criteria allowing us to adopt a definite option in
each of these cases.
We are simply looking at the pragmatic conditions under which the
exercise of the norm of objectivity takes place. It is convenient here to
take into account several possible misunderstandings of our argument.
First, from this point of view the ideal of objectivity consists in the
effort to get as close as possible to the conditions stated in our science-
fiction example; an effort to obtain a purely denotative language about
reality. This is certainly Utopian, a theoretical norm orienting the system
of social action we call 'science'. SCIENTIFIC LANGUAGE MAY BE DEFINED
AS THE CONSTANT AND UNINTERRUPTED STRUGGLE AGAINST CONNOTATION.
In the empirical sciences, this struggle manifests itself in many ways.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 71

The most important one is the effort toward a neutralization of conno-

tative meanings BY MAKING EXPLICIT THE DECISIONS THAT GENERATE
THEM. The scientific character of the construction of a descriptive and
explanatory language about reality expresses itself through the introduc-
tion of elements DENOTING THE OPERATIONS THEMSELVES WHICH HAVE
BEEN CARRIED OUT BY THE SENDER. This does not eliminate the ideological
nature of the decisions made, but neutralizes its 'ideological effect' in
communication processes. On many occasions the scientist, like the
ideologist, makes a selection of certain concepts referred to social
reality that cannot be based either on logical-methodological or empirical
criteria; but the scientist, unlike the ideologist, tries to make explicit the
very fact of having made a selection under these conditions. He decides,
for instance, to describe social structure processes in terms of class
struggle, but he warns the receiver about the fact that there exist other
possible descriptions he has decided not to choose, and that the option
involved cannot be solved, in the present state of his science, on purely
scientific grounds. As a result of this sort of DENOTATION OF THE CONNO-
TATION, the basis of the sender's operations are still ideological, but the
resulting discourse is not. The connotative level of meaning always exists,
but the 'ideological effect' derived from it only operates when the
discourse generated by the sender is presented as 'naturally' the only
possible way of talking about its subject. The 'ideological effect' dis-
appears when the selective and combinatory operations are denotated
within the language, making explicit in this way the conditions under
which the messages have been produced. To restate the difference between
science and ideology in terms of presence or absence of the 'ideological
effect' implies that this difference holds at the level of RECEPTION, i.e.,
at the level of message-consumption, and not at the level of the CONDI-
TIONS OF PRODUCTION of messages. If a set of theoretical decisions cannot,
in the present state of the social sciences, be justified by means of the
formal rules of 'scientific method', the resulting discourse has been
constructed under ideological conditions of production: it is impossible
to neutralize these conditions by 'fiat'. What may certainly BE neutralized
is its ideological consumption.16
Second, if what has been said is true, it also becomes evident that any
analysis of the problem of scientific objectivity carried on at the level of
the isolated 'scientific propositions' is entirely irrelevant. The central
issue is that of THE CONDITIONS OF PRODUCTON-CONSUMPTION of scientific
16
In this connection, see the different "readings" of the myth pointed out in Barthes,
op. c/Λ, fn. 8.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
72 ELISEO VERON

discourse within society. This issue leads us far beyond any isolated
proposition, and forces us to ask for the properties of the semantic
universe from which a given scientific text has been originated.
Third, it should be clear from what has been said that the relationship
between science and ideology has nothing to do with the issue of 'value
judgments' nor with the difference between value judgments and factual
statements. This is a typical approach to the problem that makes sense
only at the level of isolated propositions. The presence of an ideological
dimension in scientific discourse is independent from that distinction:
it stems from the system of implicit decisions underlying the construction
of scientific messages. This system supposes an evaluative dimension,
but the manifest product at the level of social communication processes
is a body of propositions that are assertive or descriptive in form. As far
as evaluation operates at the level of the conditions of message produc-
tion, it does not appear within discourse itself.
Fourth, the analysis of these problems from the viewpoint of prag-
matics liberates us from a useless discussion. Sociology of knowledge has
always had a tendency to translate the fact of the social conditioning of
knowledge into an epistemological principle. As a consequence, we face
a lot of difficulties. The most classical one is that of relativism and its
internal contradiction. If we affirm that knowledge about the social is
always socially conditioned and relative, then this very statement is
relative, and so it is not true that social knowledge is relative, et cetera.
These circles ARE vicious, and there is no utility in following them.
Epistemology defines, at a formal level, the conditions of the possibility
of knowledge; methodology, also at a formal level, states the norms
designed to fulfill these conditions. But the problem of objectivity begins
where these norms end, because it is an empirical problem. And as every-
body knows, questions of fact have to be investigated, and formal rules
do not help us to decide about empirical matters. Social scientists, once
the consensus around the formal principles has been obtained, often
make an illegitimate transposition: if the possibility of objectivity cannot
be denied, we ARE objective in our work. To raise doubts about this fact
is to be against scientific ethos. Nevertheless, all difficulties concerning
the role of ideology begin at the level of scientific PRACTICE, and therefore
this is the point where the question arises of the responsibility of social
scientists and their commitment to truth.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 73

4. THE IDEOLOGICAL USE OF THE MODEL OF SCIENTIFIC

COMPETENCE

Controversies around 'scientific social science' have seldom taken

pragmatical considerations into account. This is true — and particularly
important — of the developing nations, where great efforts devoted to the
introduction and diffusion of modern social sciences have been pursued
in recent years. In fact, the current situation allows for an ideological
interpretation. To face criticisms coming from 'the left', most 'modern'
or 'empirical' sociologists in developing countries resort to the formal
rules of scientific method: all you have to know about actual scientific
performance may be inferred from your scientific competence, According
to Germani, for instance, all problems concerning the introduction of
the theoretical orientations of American sociology into Latin American
countries may be solved by means of the rules of 'scientific method':
"... the reception of theories born in different societies or historical
periods presents itself as a problem, a problem that may be solved in a
completely satisfactory way through the general procedures of scientific
knowledge. This means that it is a purely methodological issue".17
For Galtung, the discussion of ideological problems is an expression of
'traditionalism'. His general answer to the issues involved in the develop-
ment of the social sciences in Latin America is as simple as this: "We
know of no evidence to refute a two-way model: modern society fosters
modern science, and modern science may help create modern society."18
A recent article by Cintra is another plain example of the same attitude:
opposing the strong marxist approaches within Brazilian sociology, he
does nothing but remind the reader of the principal items of an elemental
course of methodology of social research.19 At the other extreme, many
critics from 'the left' adopt the complementary position, manifesting an
implicit or explicit tendency to attack the formal rules themselves, to
question the whole analytical model of scientific procedures. It seems to
me that both positions fall into the same error: to face problems involving
SCIENTIFIC PRAXIS (i.e., involving the empirical conditions of production
of the social discourse we call science), by way of the sole analysis of
logical, epistemological, or methodological issues. The first position
17
G. Germani, La sociologia en la America Latina (Buenos Aires, EUDEPA, 1964), 4.
18
J. Galtung, "Sociocultural factors and the development of Sociology in Latin
America", Social Science Information, 5,3 (1966), 7-29.
19
A.O. Cintra, "Sociologia e ciencia", Revista Brasileira de Ciencias Sociais, 4,
1 (1966), 3-49.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
74 EL1SEO VERON

represented in Latin America by influential writers such as Germani and

Galtung, is one strictly deserving the name of SCIENTIFICISM. Scientificism
consists in the use of the normative statements referred to the formal
rules of scientific procedures, to handle problems that correspond to the
pragmatics of science. This mechanism usually emerges in situations of
strong ideological cleavage. In fact, MOST OF THE PROBLEMS REFERRED
TO THE RELATIONSHIP BETWEEN THE PRACTICE OF THE SOCIAL SCIENCES IN
HIGHLY DEVELOPED COUNTRIES AND IN UNDERDEVELOPED ONES, AND TO
THE INTRODUCTION OF THE SOCIAL SCIENCES IN THE LATTER, ARE UNDECID-
ABLE IN TERMS OF THE FORMAL NORMS OF SCIENTIFIC PROCEDURE, AND
THEREFORE IMPLY IDEOLOGICAL DECISION-RULES.
Within this framework, PSYCHOLOGISM becomes the 'natural' com-
panion of scientificism: ideology intrudes into scientific activity as a
result of the influence of the scientist's 'subjectivity' or of 'emotional'
processes. In this light, ideology appears as an alien factor within
scientific praxis, as a disruptive element coming from Outside'. Long
ago, Popper pointed out the fundamental mistake of the traditional
sociology of knowledge, which consisted in referring the problem of
scientific objectivity to the problem of the subjectivity of the scientist.20
From the point of view of communication theory, ideology is a level of
meaning, and this implies that it is A STRUCTURAL CONDITION OF PRO-
DUCTION OF MESSAGES WITHIN A HUMAN LANGUAGE SYSTEM, INCLUDING
SCIENTIFIC COMMUNICATION.

5. THE TWO STEPS OF PRAGMATIC ANALYSIS

Let us summarize our central argument. We think that the pragmatic

approach is the only way of stating, in an adequate and complete way,
the problem of scientific objectivity and of the relationships between
science and ideology. It is adequate since it allows us to locate the level
at which the problem of objectivity actually appears: science viewed as
an empirical system of social activity oriented toward knowledge. At the
same time, it allows us to unmask what is nothing else but an ideological
operation: trying to solve an empirical problem with formal criteria,
those stemming from epistemological and methodological principles.
It is complete, because it integrates the other levels of analysis of scientific
activity within the global context of functioning of science. It would
20
K. Popper, The Open Society and Its Enemies (Princeton, N.J., Princeton Uni-
versity Press), Chapter 23.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
IDEOLOGY AND SOCIAL SCIENCES 75

then be a serious mistake to consider that the pragmatic analysis of

scientific communication is in some way opposed to the viewpoints and
approaches derived from modern epistemology and methodology of the
social sciences. The study of the formal conditions of scientific knowledge
and that of the norms regulating their exercise are two indispensable
levels of analysis of science. But the third level (that of pragmatics)
integrates the other two, and is the only one giving a complete picture
of scientific knowledge as ä process of production within society.
As we pointed out before, the analysis of the rules of procedure, of the
techniques and of the conceptual bodies existing in a discipline at a given
moment as fields of alternatives within which scientific decisions are
moving, constitutes the starting point of pragmatics, which comprises
two large aspects or steps: one of description, the other of explanation.
First, there is the task of 'mapping' these fields of alternatives (theo-
retical, methodological, technical) at a given moment of development
of a discipline: which version of methodology of science is predominantly
used; what techniques of data gathering are preferably applied to what
problems; which are the structural properties of the conceptual frame-
works used to locate the different aspects of social phenomena, etc.
Second, once this picture has been drawn, we may ask why it presents
such characteristics. This leads us by necessity to the social context:
relationship of intellectual and scientific elites with the dominating
classes; state and dynamics of class conflict within the country under
study and location of that country in the international system. This type
of analysis must include the study of the mechanisms mediating the
relationship between the sociological variables and the products of
scientific knowledge: organizational characteristics of academic institu-
tions; systems of distribution of power among different academic groups
and of allocation of financial resources; personnel selection and training
rules; rules of control and diffusion of the produced knowledge, etc.21
It goes without saying that such a research program cannot be 'neutral'.
So we find here again, on a new level, the same problems we have dis-
cussed before, and the same criteria have to be applied. The development
of explanatory models in pragmatics demands theoretical and metho-
dological decisions, and the selection of particular research techniques.
Since pragmatics as we have presented it is in fact a SOCIAL science of
scientific communication, it could be said that our argument implies a
21
An analysis of the current situation of social sciences in Latin America may be
found in E. Veron, "Ideologie et production de connaissances sociologiques en
Amorique Latine", VHomme et la societt (forthcoming).

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM
76 ELISEO VERON

contradiction or contains a. vicious circle. This objection is purely

formal and deserves, therefore, a formal answer: there is no contradiction
or circularity inasmuch as it is clearly established that pragmatics of
science is, with regard to scientific language, a metalanguage, a level of
description of a higher degree of complexity. The distinction between
levels of language eliminates the possible logical or formal contra-
dictions. In return, I deem it correct to say that this way of stating the
problem contains a regression to infinity, since one can imagine a
pragmatic analysis of pragmatics of science, and so on indefinitely.
This is a particular illustration of a known principle: the hierarchy
between levels of language implies an Open system' that cannot be
axiomatically closed. And a regression ad infinitum is something that
should not trouble us: it is certainly a long way, but free of contradiction.

Eliseo Veron was professor of Systematic Sociology at the University of Buenos

Aires, Argentina, until 1966. At present he is a full-time researcher at the Torcuato
Di Telia Institute, Center for Social Research, where he is Director of a Research
Program on Social Communication. He worked from 1961 to 1963 under a post-
graduate Fellowship at the Laboratoire d'Anthropologie Sociale in Paris, and from
1965 to 1968 was Editor-in-Chief of the Latin American Journal of Sociology. He has
published three books: Conducta, estructura y comunicacion [Behavior, structure and
communication]; Comunicacion y neurosis [Communication and neurosis]; and, as
editor, Lenguaje y comunicacion social [Language and social communication]; and
many articles in social science journals. He translated into Spanish and wrote a
preface to L£vi-Strauss's Structural Anthropology, and is now in charge of the Spanish
edition of Communications, the journal of the Centre d'fitudes des Communications
de Masse (E.P.H.E.). He is currently working in the field of semiolotics applied to
mass-communication.

Brought to you by | The University of York

Authenticated
Download Date | 2/18/16 3:03 PM