semiotic

1/ When we talk of 'scientific research in social sciences', and when we talk of 'scientific research in natural sciences' we talk in both cases of 'scientific research'.

2/ This just stated, it seems intuitively true that together with the existence of scientific research work, we assert the possible existence of scientific work which is not research work, research work which is not scientific, and of course, work which is neither scientific nor research.

3/ It is sometimes questioned or argued whether it is correct or not, when considering 'social sciences', to talk of 'science' and of 'scientific research'. Such a doubt or a discussion implies that a certain kind of work may be carried out within the field of social sciences, though being of a nonscientific nature, or while being neither scientific nor research.

4/ When we state the problem as concerning the so called 'social sciences', we are assuming that social phenomena may be subject matter of scientific knowledge. Otherwise, we would have to refer to 'social disciplines or doctrines'.

5/ The argument that follows will only interest those who share (or are ready to discuss honestly) the assumption of the existence of social sciences. It would not be suitable herein, to take the critical treatment of the problem as far back as to the primal reasons that justify such an assumption (without overlooking the fact that in some academic circles it still is important, though nowadays of little interest). Therefore, it will be admitted (at least for the time being) that there is the possibility of scientific knowledge of social problems. the set of languages that make such knowledge possible will be herein called 'social sciences' (i)(ii).

6/ According to this assumption, there exists scientific research in social sciences, as well as other sorts of tasks which are not research though being scientific. The set of tasks which are nonscientific and, of research activities which are nonscientific will not be within the scope of the above stated field. Of course, task that is neither scientific nor research will also be excluded.

7/ This paper will only deal with some aspects of scientific research in the field of social sciences.

II. The Meaning of Social Phenomena

8/ The subject matter of social sciences is the universe of social phenomena.9/ 'Social' herein means the representation/interpretation, present or historical, of some phenomenona, that obtains in a certain community, including both the so called 'cultural' and 'natural' phenomena. Thus, the quality of being social does not concern a certain field of occurrence of the phenomenon but a certain operation which causes it (in what concerns the way it appears and appears to be) (iii)(iv).

10/ 'Representation' herein is understood as the specific perceptual identification (sensorial or imaginary) of certain forms (and their interrelationship), which a certain phenomenon takes, in terms of its possible interpretation, at a certain moment in a certain society (v)(vi)(vii).

11/ 'Interpretation' herein is understood as the conceptual assignment of a given meaning to a certain phenomenon in terms of its possible representation, at a certain moment in a certain society (viii).

12/ 'Representation' and 'interpretation' are therefore recursively defined. Such terms will be herein used indistinctly or together to specify the quality of the information in social sciences, but they will be differently handled according to the approach to the pertinent subject matter (ix).

13/ To say that 'social phenomena are the subject matter of social sciences' is like saying that 'social sciences study the forms the social representation/interpretation of any sort of phenomena takes' (x).

14/ Herein, we mean by 'investigation' in the field of social sciences the process of elaboration of an explanation about how, and/or why, and/or with what result, and/or from what precedent, a certain phenomenon is represented/interpreted in a certain way at a certain moment in a certain society (xi).

15/ There exists, for instance, a legal way of representing/ interpreting any sort of phenomena (from an earthquake to the publishing of a book), and there exist other ways (psychological, sociological, historical, anthropological, communicational, linguistic, etc.) of representing/interpreting the same phenomena. Each of these ways constitutes the specific subject matter of each one of the social sciences.

16/ For research work in the field of social sciences to be scientific, the process of elaboration of the above mentioned explanation must be carried out using correctly and properly (that is, according to the appropriate syntactic and semantic rules) the specific language of the corresponding science.

17/ The language of social sciences accounts for the different (social) ways of representing/interpreting any kind of phenomena whatsoever. This (social) elaboration of phenomena is already a language: the one through which the members of a certain society intuitively interpret the possible representation of the phenomena of their environment. Thus, the language of social sciences is a metalanguage (xii)(xiii)(xiv).

18/ Scientific research work in the field of social sciences needs previously a certain set of intermediate operations that would establish the bidirectional correspondence between the language (natural-intuitive) which produces the social representation/ interpretation of a certain phenomenon and the metalanguage (theoretical-critical) which produces its scientific explanation. These intermediate operations constitute the research methodology in social sciences (which is, therefore, a second degree metalanguage) (xv).

19/ From now onwards, this work will solely deal with some aspects that concern the metalinguistic characteristics of scientific research methodology in the field of social sciences.

III. Data, Information and Hypothesis

20/ A social science researcher must have clear in mind certain aspects concerning the initial stage of his work: the compilation of data and information (xvi).

21/ In principle, it can be asserted that his data are not the phenomena but the social discourses about the phenomena (xvii).

22/ It can be asserted, in principle too, that his information is not the data but the representations/interpretations which can be identified in them (that is, in social discourse(s)) (xviii)(xix)(xx)(xxi).

23/ The social representation of a phenomenon is the perceptual correlate of the social interpretation of such a phenomenon, and both are materialized in social discourse (verbal, graphic, objectual, behavioral, musical, etc.) which is not only its physical support but, also, the instrument that makes possible its production.

24/ In this work, it is called 'social discourse' the actual set of constructions circulating in a society, that are used efficiently for the effective production and/or reproduction of perceptual representations and of conceptual or valuation interpretations (xxii).

25/ 'Representation', 'interpretation', and 'discourse' are terms which herein name the (numerically infinite) sets of representations, interpretations and discourses.

26/ In sum, the data for research work in social sciences is, therefore, the social discourses in which the interpretations assigned to the representations of certain phenomena given at a certain time of a certain society are materialized.

27/ The interpretation given to the representation of a certain phenomenon, as it is embodied in a certain social discourse, is the basis information for research on social sciences.

28/ The direct and personal perception of the phenomenon the researcher may have is only one of the many socially interpreted representations.

29/ When the researcher includes as data his personal perception of the phenomenon, he is only adding another social discourse. As such, he can do so. Otherwise, this may acquire in research work the value of a privileged discourse (since he contrasts the remaining statements on the phenomenon under study with his own observation, holding the latter as the most reliable guarantee of objectivity).

30/ The presumption of a personal and direct perception of the phenomenon prevents him from seeing (by making him lose sight of it) the interpretative/representative character of any human perception (that is, the character of already being interpreted, which all phenomenon, being representable or perceivable, has). Thus the character of subject matter of social knowledge that the phenomenon under study should have is banished, this knowledge being only possible if the phenomenon has already been interpreted in one (or several) social discourse(s) and in terms of one (or several) of its socially available representations.

31/ The assertion that what the researcher needs to identify as basic information consists of the set (or subset, whose relevance he will have to justify) of present or historical representations/ interpretations of the phenomenon under study, is not a mere rule of methodological procedure but a real cognitive requirement (at the present stage of the cognitive science). Any identification or description of a phenomenon stems from a certain social system of representation/interpretation. The system of values from which the phenomenon is identified, described and interpreted is implicit in the presumption that you can grasp the phenomenon in itself and objectively (versus the Kantian rejection of 'das Dinge an sich'). In this case, what should be the subject of research (the intended explanation) is taken as already applied to the phenomenon. If we know what the phenomenon is (in as much as it is recognizable by its own features), then it is not subject matter for cognition but for recognition (as an intuitive assertion of such features). But there is no original or uncontaminated intuition (that is, all intuitions are historically conditioned). Thus, to take the phenomenon as a positivistic and non-interpreted datum means to take it the way some (non explicit) interpretation makes its representation possible. The intended interpretation would only develop the set of possible variations within a certain system of values (xxiii).

32/ This is, in fact, the way in which each member of a certain society represents and interprets the phenomena that surround him. The scientific researcher, in his specific task, does not behave as a native user of the systems of representation/ interpretation in force in his society. He is the user of a scientific language (a metalanguage) through which he explains the way native languages function in the production of such representations/interpretations.

33/ As there is no immediate access to the phenomenon in itself, there is not either immediate access to the interpretation or representation (in itself) of such phenomenon; this is not a mere rule of methodological procedure either, but an answer to the requirements of logical reasoning (at the present stage of the science of Logic); interpretation and representation are inferences (whose production must be proved); their only observable entity is the social discourse (xxiv).

34/ Wen producing the representations and interpretations that are possible at a certain moment in a certain society, the social discourse materializes them; but when using the cognitive operations available at a certain moment in a certain society for the production of representations and interpretations, the social discourse also materializes these operations.

35/ Understood as such, the social discourse is the observable entity (the datum) on which the researcher in social sciences works. In this way, he establishes, through the pertinent analytical operations, the features of representation/interpretation (the information) with which he will develop the explanation about the meaning a certain phenomenon has at a certain moment in a certain society.

36/ When compiling data, the researcher, explicitly (the best way) or implicitly, states a hypothesis: that these data are adequate, necessary and sufficient to develop the explanation he intends to give. This is, therefore, a risk the researcher assumes: to demonstrate throughout his work such adequacy, necessity, and sufficiency. The sum of these features is herein called: data 'pertinence' (xxv).

37/ These three features are not, therefore, inherent in the phenomenon but in the social discourse.

38/ Consequently, the compilation of data cannot be evaluated in relation to the phenomenon (this would imply perverting the concepts of 'objectivity' or 'reality'), but to the representation/ interpretation whose production, in connection with a certain phenomenon, the researcher intends to explain.

39/ The objectivity and reality of a certain representation/ interpretation of a given phenomenon, in as much as it is actually produced by and materialized in a certain social discourse at a certain moment in a certain society, will be herein called, its 'force' (xxvi).

40/ This state of affairs is more risk taking for the researcher, as the hypothesis on the pertinence of the data he compiles depends on another previous hypothesis, connected to the existence of the information that he will have to identify through analytical operations. Such a hypothesis depends, in turn, on the probability of an initial hypothesis on the subject matter of the explanation he proposes for the phenomenon under consideration.

41/ That is, the researcher states (in 3rd place) the (working) hypothesis that the data he compiles is pertinent, as a result of having stated (in second place) the (methodological) hypothesis that certain analytical operations will enable him to identify (in such data) the prevailing relations used for the representation/ interpretation of a certain phenomenon, in terms, consequence, in its turn, of the (theoretical) hypothesis stated (in first place) on the provability of the explanation in question.

42/ 'Provability' of an explanation herein means the possibility of proving that the meaning of a phenomenon is built on the concrete efectiveness of certain representations/interpretations, at a certain moment in a certain society (xxvii).

43/ The investigation may show that the (3rd) working hypothesis was false, say, that the compiled social discourses were not pertinent to identify the (representation/interpretation) relationships capable of acheiving the intended explanation.

44/ The (2nd) methodological hypothesis may be false as well, when it is proved that certain representations/interpretations actually present in the analyzed social discourse are not the prevailing representations/interpretations, with which a certain society builds on the meaning of a certain phenomenon at a certain moment.

45/ Obviously, the (1st) theoretical hypothesis may be false as well, when it is proved that none of the representations/interpretations in force in a certain society manage to prove the (assumed) meaningful content attributed by the researcher to a certain phenomenon.

46/ In the first two cases, the investigation goes as far back as to the initial stage, making it necessary to plan a new compilation of data, in the first case, and new analytical operations, in the second one. Failure in one of these hypotheses does not necessarily imply failure in the other, but there may be joint failure as well; that is to say, the representations/interpretations may be the existing ones to give meaning to a certain phenomenon, but they are not embodied in the compiled social discourses; or the social discourses contain them but the analytical operations are not effective enough to be identified. Of course, both failures may concur and in this way the theoretical hypothesis on the meaningfulness of a certain phenomenon is still a mere assumption , provable but not yet proven (xxviii).

47/ In the third case, that of failure in the theoretical hypothesis (which can only be determined through an adequate, necessary, and sufficient compilation of the social discourse, and through the application of effective analytical operations for the identification of the representations/interpretations socially in force, but which do not lead to prove the intended explanation), it is necessary to state a new explanatory hypothesis. Usually, this new hypothesis will have already been proved in the praxis of research. The content of its conclusion would contradict the contents of the initial hypothesis, but investigation would be valid simply stating such a contradiction explicitly or even (in as much as science is an act of communication) substituting, in the final report, the contents of the failed initial hypothesis and stating, in its place, the contents of the conclusion attained (xxix)(xxx).

48/ However, every scientific research work results in knowledge; in the above mentioned cases (failure in the hypothesis), the lack of pertinence of data, and/or the non-operation of the information, and/or the non-provability of the explanation would have been established (at least temporarily). These exclusions imply an increase of knowledge.

49/ In sum, there is no valid data in itself but useful data forágiven purposes. There are no analytical operations which are effective in themselves but effective forágiven purposes. And, of course, there are no explanations necessarily provable but just possible.

IV. Analytical Operations

50/ Hence, the researcher in social sciences also needs to bear in mind certain aspects of the second one of the main stages which his task consists of: analysis of the information.

51/ 'Analysis of the information' herein means the application of a set of technical operations through which the information produced/reproduced in the compiled social discourse (object language) is identified, contrasted, and transformed (2nd metalanguage) according to the specific rules of each one of the social sciences (1st metalanguage) (xxxi).

52/ This set of technical operations does not differ, as regards its nature, from those through which the available information is interpreted by each member of a given society (xxxii).

53/ This set of technical operations differs essentially, as regards the constraints imposed on its application, from those through which the available information is interpreted by each member of a given society.

54/ The confusion or, at least, the non-distinct delimitation between this proximity in the nature of these operations, and this divergence in the constraints of its application, keeps the scientific status of the knowledge on social phenomena, and of its corresponding research, still in debate.

55/ The non formalization or the partial formalization of the specific languages in the different social sciences also contributes to keep the debate still alive. 'Formalization' herein means the existence of explicit definitions of the theoretical and methodological terms used in such sciences. In addition, of course, the set of such definitions will have to fulfil the general logical constraints of being complete, consistent, and decidable.

56/ The first obstacle: the inadequate differentiation between the nature of the operations by which we make interpretations (shared by all human beings), and the constraints imposed on the application of such operations (specific to the scientific status of the interpretation), is easy to avoid (at least in theory): the researcher in social sciences cannot allow himself to give an intuitive interpretation of the information under analysis. 'Intuitive interpretation' herein means the one which is attained using, in a non-critical way, the existing social languages (xxxiii).

57/ This does not mean that scientific interpretation cannot coincide with intuitive interpretation. It is not the contents but the procedures through which the interpretation is made what must differ. What makes an interpretation scientific is the procedures, regardless of the coincidence between the obtained result and intuition (xxxiv).

58/ The second obstacle: the fact that the language of social sciences is in process of formalization (since only partially formalized theoretical systems exist in this field) provides the researcher with propitious space for scientific innovation, with greater possibilities (and more risks) than those existing in the field of natural sciences (xxxv)(xxxvi)(xxxvii).

59/ In this way, the task the researcher in social sciences has the moment previous to the analysis and keeping in mind its application, lies in the formalization of his reference metalanguage (that is, the statement of the definitions of theoretical and methodological terms, by means of which the researcher will analytically take part in the compiled social discourse) (xxxviii).

60/ The outcome of this formalization (usually called 'theoretical frame', which does not consist of the statement of the total number of current theories which can eventually affect his subject matter, but of the critical statement of the available theoretical language, in the corresponding state of science, and which can affect his subject matter) will include: a) explicit definitions (of the terms naming concepts and operations) arising from existing theories; from only one theory, or from the synthesis of several (paying special attention, in the latter case, to the problems arising from the commensurability or incommensurability of theories), and b) own (that is, stated by the researcher himself) explicit definitions (of the terms naming concepts and operations) in those aspects (concerning the subject matter of his research) of which the available theories do not give account (or which have not been formalized in them, or whose formalization does not satisfy the researcher).

61/ In any case, the researcher will take special care to keep a logical coherence into the set of definitions he uses, according to the above mentioned criteria: completeness, consistency, and decidability.

62/ This legitimate attitude, susceptible of being included in a theoretical 'constructivism', has helped, due to the deviating features of its use, to discredit social sciences; at least this is so in academic circles where theoretical structures are wholly formalized, and even stated in a symbolic language wich limits even more the margin of terminological ambiguity. What is to be censured is the lack of explicitness of the definitions of the terms to be used, or the incompleteness, inconsistency, and undecidability of such terminology, and hence the ad hoc modification of these definitions (or the change in the meaning of the non-defined terms), to make them fit the development of the research work. This ambiguous or at least soft treatment of the social phenomena has moved the outcome of social sciences closer to those of literature, where the terminological ambiguity or polysemy is the procedure for aesthetic creation (explicitly acknowledging the effectiveness of aesthetics in the production of knowledge). Theoretical constructivism claimed here as legitimate, demands a hard operational and terminological treatment (as to specific to the scientific mode of production of knowledge). It is important also in this case to assert this differential criterion, though its development becomes less interesting (xxxix).

63/ Once the researcher has at his disposal (or builds on) a formalization of his theoretical terms (the 1st degree metalanguage of the pertinent social science) in what concerns the specification of the subject matter he intends to do research on, he goes on to deal with the identification, definition, and functional description of each of the analytical operations (the 2nd degree metalanguage, or methodology strictu sensu) which he considers necessary to apply to the compiled social discourse.

64/ By 'analytical operations' herein we mean the set of procedural rules that frame the specific way in which the researcher works with the social discourse. They aim at proving the existence, in such social discourse, of certain observable relations and their productivity as regards the representations/interpretations whose existence and effectiveness have been stated in the theoretical hypotheses.

65/ It is worth noting that neither the theoretical hypothesis (nor, let alone, the analytical operations) are a priori models according to which the relationships observable in the social discourse would be interpreted.

66/ This does not exclude the epistemological dependance of such hypotheses and operations on the scientific historical juncture at which the researcher works. A pure, absolute, and nonhistorical scientific reasoning does not exist (xl).

67/ The theoretical hypotheses state a supposition about the force of certain representations/interpretations which are effectively produced by the social discourse, and are effective producers of the meaning of a certain social phenomenon. They do not build on nor rebuild on the social discourse but they conjecture that social discourse as producer of such representations/ interpretations may be effective (xli).

68/ Analytical operations, being procedural rules, have the heuristic value of questions or interrogations about the way representations/ interpretations are produced. The answers to such inquiries will be found (if successful) in the social discourse and depending on the actual mode of production effectively shown by such discourse. Therefore, the function of analytical operations consists of building models (whose validity will be restricted to the analyzed information) and not of applying or verifying them (xlii)(xliii).

69/ 'Built model' herein means a possible interpretation of a given social discourse, obtained by means of the application of a certain set of analytical operations, originating with a supposed interpretation previously stated in the theoretical hypothesis or hypotheses. The existence and the specific features of such a model are an outcome which will only be cognoscible a posteriori of the investigation.

70/ When knowledge is so acquired, it is rigorous, merely possible, and virtually fallible; but it is historically plausible. These are the fundamental features of scientific truth; which only is (must be) consistent with the theoretical system that originates it (xliv)(xlv)(xlvi).

71/ The effectiveness of such knowledge (rigorous, possible, fallible, and plausible) as an explanation of the operational meaning of a certain social phenomenon in a certain society, is the basis for its practical usefulness (in what concerns the possibility of political performance in the social juncture), and sets (at least implicitly) the historical time limits of its validity (historicity of scientific knowledge).

72/ The analytical operations constitute herein the specific device for the (scientific) production of such (scientific) knowledge.

73/ Hence, analytical operations are herein characterized as being formal, independent of the theoretical hypotheses, and transferable.

74/ An analytical operation is herein meant to be 'formal' in so far as it does not imply the contents of its result.

75/ An analytical operation is, herein, considered to be 'independent of the theoretical hypotheses' in so far as, with reference to the social discourse it is applied, it can validate or invalidate them. And also in so far as it keeps its usefulness to prove other theoretical hypotheses, different or even contradictory to the original ones (when applied to a different social discourse).

76/ An analytical operation is, herein, said to be 'transferable' in so far as another researcher can learn to use it, and will (when using it) obtain the same results if he applies it to the same social discourse.

77/ Herein, analytical operations , as a result of their procedural features, may be: identifying, contrastive, and transformational. This enumeration only states the basic analytical operations in social sciences. Each of the remaining specific operations of each particular social science will be included in one or another of these three main groups of basic operations, or will combine the features that define them (xlvii).

78/ 'Identifying analytical operation' herein refers to the one which separates and links perceptual marks actually present in a certain social discourse (xlviii).

79/ Its application depends on the theoretical hypothesis (or assumption) that asserts that certain marks (and not any others) produce the representations/interpretations with which a certain society attributes a certain meaning to a certain phenomenon at a certain moment.

80/ Its success explains the material mode of production of the meaning of a certain phenomenon: syntactic component of meaning.

81/ 'Contrastive analytical operation' herein refers to the one which relates the perceptual marks identified in a certain social discourse to other perceptual marks identified in a different social discourse, synchronic with the former (xlix).

82/ Its application depends on the theoretical hypothesis (or assumption) which asserts that the contrast between the marks of such (synchronic) contrasted social discourses gives, the representations/ interpretations produced or reproduced in one of such discourses (which we will call the substituting one), their specific differential meaning.

83/ Its success explains the differential (and competitive) mode in which each social discourse produces the meaning of a certain phenomenon, at a certain moment in a certain society: semantic component of meaning.

84/ 'Transformational analytical operation' herein refers to the one which relates the pairs of marks contrasted in two (sets of) different social discourses belonging to a certain moment (a certain present time) of a certain society, to other pairs of marks contrasted in two other (sets of) different social discourses occuring at some certain other moment (previous or subsequent to such a present time) of the same society (l).

85/ Its application depends on the theoretical hypothesis (or assumption) that asserts that the possible transformation between the specific differential meaning of a certain phenomenon occuring at a certain given present time, and the specific differential meaning of such a phenomenon at anotherágiven moment (previous or subsequent to such a present time): a) has not taken place, so the meaning of the phenomenon is still the same; or b) has taken place, so the differential meaning is not only different but it is such that one of the different meanings overcomes (or is overcome by) the contradictions inherent in the other.

86/ Only when the proof of such overcoming is successful, will the researcher be able to assert he has achieved the entire explanation of the meaning of a certain phenomenon, since, apart from the contrasted meaning obtained in relation to a certain moment, he will have established the historical mode of such a meaning: dialectical component of meaning.

87/ 'Overcoming' herein means the transformation of the peculiar features of the basic initial contradiction (inherent in the fact that the meaning of a certain phenomenon occurs in the social discourse and not in the phenomenon in itself), into another subsequent contradiction of peculiar features (whose structure will only render it self conscious with the use, throughout time, of the social discourse that contains it), which will, in its turn, have to be overcome (li).

88/ The successive overcoming of this sequence of contradictions constitutes the historical framework of social phenomena. Rebuilding it, till the moment the researcher starts his work, or explaining the phenomenon in terms of the historical framework of the social discourses that have given such a phenomenon its meaning, constitutes the subject matter of research in the field of social sciences.

89/ What has been developed in the preceding paragraphs states basic methodological guidelines for research in social sciences, arising from the critical reflection on the possibility of a hard semiotics, modelled in conformity with the logical structure of the 'possible semiotic worlds'.

NOTES

(i/#5) The suggestion that the problem of the actually scientific quality of the social sciences 'is still important' recognizes its political trascendence, as a claim for its own academic space. The realization, however,that it is now 'of little interest' implies the theoretical overcoming of such problem, due both to the widespread rejection, in natural sciences, of the logics of justification of classical epistemology -empirical or rationalist- (I. Lakatos, 1981: 176) and to the widespread concern to establish 'the form of positivity' of social sciences (M. Foucault, 1966: 363).

(ii/#5) Defining science as a language does not constitute a specific perspective of social sciences, but one that is pertinent to science in general. N. Bohr's criterion is followed in this respect : 'We are suspended in language in such a way that we cannot say what is up and what is down' (quoted by J. Bub, 1974: 404).

(iii/#9) This concept of phenomenon, postulated as specific to social sciences excludes the Husserlian possibility of knowledge of the phenomenon in itself and adopts Kant's tradition again, in as much as, in the phenomenon, objects and the way we consider they are, are always deemed to be something real; but as long as the way they are depends only on the mode of intuition of the subject in his relation with the given object, this object is not the same as phenomenon as what it is as object in itself (Kant, Critics of Pure Reason: Trascendental Aesthetic, 8, III). Likewise, C. S. Peirce defines the phenomenon (or 'phaneron') as 'the collective total of all that is in any way or in any sense present to the mind, quite regardless of whether it corresponds to any real thing or not' (Peirce: 1.284). Phenomenalism which Peirce assumes with total awareness of being 'the phenomenalism of Kant, and not that of Hume. Indeed, what Kant called his Copernican step was precisely the passage from the nominalist to a realistic view of reality. It was the essence of his philosophy to regard the real object as determined by the mind' (8.15). This is an idealistic remnant (and much to Peirce's displeasure, also a Hegelian one) that in the constructive view of the phenomenon is excluded, while reasoning is determined by that discourse carried out through the social use of the language. Cf., for example, the explicit materialism of M. Pˆcheux, who is hopeful about Mel'cuk and Zolkovskij's theoretical project and quotes their definition of 'sense': 'Sense means what is common to the various utterances recognized and used by speakers as synonyms, or in short, sense is the invariant of synonymous transformations (paraphrase)' (M. Pˆcheux, 1975: 268).

(iv/#9) In this sense, also 'Ren‚ Thom's opinion is that "every science is the study of a phenomenology"' (quoted by R. Marty, 1986: 341).

(v/#10) This calls for awareness not to make the stimulus mistake or thing mistake or object mistake, that is, the assumption that the mental perception of a certain thing is identical to all or some of the objective properties of that thing (R. Arnheim, 1971: 103)

(vi/#10) The concept of representation defined in this paragraph synthesizes the two notions reciprocally connected, which, in terms of Janik and Toulmin, were not clearly distinguished in Kant's and Schopenhauer's time, and which have been confused till now: in one of its senses, the term had a 'sensorial or perceptual' use -so was the case in Helmholtz's physiological vision or in Mach's psychological vision- which connected it with Lock and Hume's empirical phylosopies. In its other sense it had a more 'public or linguistic' use, as in Hertz's mechanics, analogous to that of the expression 'graphic representation' of present-day physics (A. Janik and S. Toulmin, 1974: 166). This synthesis 'constitutes the founding relation of epistemology, the relation between knowledge and representation, taking into account the new concepts provided to scientific research by the new techniques of the representation of knowledge' (J.-L. Le Moigne, 1986: 52). Its antecedents come from C.S. Peirce: '(...) our perceptual judgements are the first premises of all our reasoning (...). All our other judgements are so many theories whose only justification is that they have been and will be born out by perceptual judgements' (5.116). See: 'perceptual judgements', not 'perceptions', in the same sense as we are herein talking of 'representations'.

(vii/#10) It is also important to keep the substitutional character of every representation. 'A representation is that character of a thing by virtue of which, for the production of a certain mental effect, it may stand in place of another thing' (Peirce: 1.564). Thus, the representation of a phenomenon is not the phenomenon in itself, but something that stands in its place, in as long as a certain community accepts it or establishes it as such.

(viii/#11) The use of the term 'interpretation' herein is quite distant from the characteristic approach of hermeneutics, which explores the interpretative effectiveness either in the subject: 'the interpretative discourse depends on the interpretative subject, qualified by some preliminary knowledge and an explanatory competence which make his involvement in the search of sense possible' (S. Alexandrescu, 1979: 217), or in the comment, 'that transmits some kind of knowledge whose pretensive to what is interpretative proper: 'the comment interprets the statement taken as object' (L. Panier, 1979:240). Hermeneutics deals with 'what is unfinished, the infinity of interpretations' and hence 'it is necessary to understand (...) that hermeneutics and semiology are two unfortunate enemies. In fact, a kind of hermeneutics which falls back upon a kind of semiology must believe in the absolute existence of signs', something which contradicts the time of (hermeneutic) interpretation, that is circular (M. Foucault, 1964: 8). This hard sense of semiotics is the one developed in this work as a methodological basis which includes the concept of interpretation. It is also the analytical sense in which R. H. Thomason writes in his 'Introduction' to R. Montague's work that instead of considering past time as acting on the extensional values of the formulae, we will place its influence on intensions, being these, as extensions, possible denotations built on an interpretative structure' (1977: 42). Thus, some very general considerations on the nature and use of analytical operations (as for interpretative structures) are herein stated to help the researcher identify the possible (intensional and extensional) denotations of social discourses which produce the significance of the phenomena under consideration. This is also the sense in which M. Bunge characterizes interpretation (with the only objection that, as this work refers to the field of social sciences, the interpretation of an event is considered to be always mediated by the interpretation of a sign, in the broad sense of text or discourse, being this mediation the specific object of investigation): 'We interpret a fact when we explain it and we interpret an artificial sign (symbol) when we find out or stipulate what it means in a certain context. And an artificial sign means -if it means something- what it represents, that is, its designatum (...). In particular, a unit sign or term is signficant if it designates a non-empty set. The designation is non-ambiguous if the set is a unit-set. And a sentence will be significant if it represents a set of propositions. The sentence will be ambiguous unless it represents only one proposition and it will be non-sense if it does not represent any proposition' (M. Bunge, 1973: 161). Later on, he refers to the 'interpretative explanation' which he differentiates from the 'subsumptive explanations', asserting that 'only representational theories, those theories which aim at representing the modus operandi of their correlates, can offer deeper explanations. We will call them interpretative explanations' (Ibid: 585). The difference with Bunge lies in considering that in social sciences the 'modus operandi' on which such representation/interpretation falls back and which therefore constitutes the object of the pertinent research, consists of the material productivity of the natural language (of the empirical social way which it may manifest itself) for the production of the significance of the phenomena. Natural language (with its plurality inherent in its social character) plays, as regards social phenomena, a role similar (but in such a multiplied and twisted way that it is essential the delimitation of scope which the researcher imposes on himself for the sake of its study) to that played by scientific theories in connection with the facts they are applied to. As regards the latter, P. K. Feyerabend says in his classical article: 'Facts and theories are much more intimately connected than it is admitted by the autonomy principle. Not only is the description of every single fact dependent on some theory (which may, of course, be very different from the theory to be tested); also, there are facts which cannot be unearthed but with the help of alternatives to the theory to be tested, and which become of no use as soon as such alternatives are excluded. This suggests that the methodological unit to which we must refer when discussing questions of test and empirical content is constituted by a whole set of partly overlapping, factually adequate, but mutually inconsistent theories' (P. K. Feyerabend, 1970: 330). Just as the epistemologic reflection has a metalinguistic level in relation to the scientific (natural) theories under consideration (and here becomes relevant Feyerabend's reflection on the existence of 'facts which cannot be unearthed' except by taking into account the multiplicity of theories, or at least, their alternatives), social sciences have as well, being formulated theories, a metalinguistic level for being part of the multitude of social languages; metalanguages without which no phenomenon could be unearthed (that is to say, the intepretation of the effectiveness of such languages for the production of the significance of a given phenomenon could not be achieved).

(ix/#12) Although the topic of recursiveness or recurrence is appealing due to its usefulness for the knowledge of the processes of production of knowledge (as from the formulation of Peano's fifth postulate; see for this J. A. Magari¤os de Morentin, 1992) it is herein mentioned, only as a metatheoretical reflection on the definitions in the previous paragraphs 10-11. The purpose has been to bind together, through a conceptual necessity, the social operations of representation and interpretation, establishing the inability to define one without having to resort to the other, but keeping their independence (which is not the same as the comments made by N. R. Hanson, whose interest is to deny the duality of seeing and interpreting: we dare say that interpretation is vision; 1977.However, Hanson also makes some interesting observations on the difference between 'see how' and 'see what', something which makes it possible to bring his ideas of seeing or vision closer to the one herein stated of representing or representation). The problem which recurrent definitions bring about, in the sense they are herein stated, is closely connected with those definitions which H. Poincar‚ called 'unpredicative definitions', and which Russell called 'the principle of the vicious circle', in the critique of G. Frege, when stating his theory of logical types (W. S. Hatcher, 1968: 155s).

(x/#13) There's a subtle line of difference here between natural and social sciences. Of course, in both cases, we are talking about sciences, hence the difference is not stated with regard to this common concept. The non-immediateness of their empirical object is also common to both, but the difference lies in the instrument of mediation. In sciences whose object is the social phenomena, the mediation upon which scientific reasoning works is, as herein supported, the social discourse. In sciences whose object is the natural phenomena, the mediation is given by the historical scientific discourse (as an antecedent to the one of the researcher who acts at a certain moment and perceives the object through the prior discourse). However, this must be understood as predominance and not in an absolute way. The subject matter of natural sciences is also mediated (not simply by the specific historical scientific discourse) by the social discourse which at any rate attributes to them a contingent relevance and sometimes suggests ethical and aesthetical ways for their representation / interpretation. These representation / interpretation ways are experienced by the scientific reasoning (when they are) as anticipations, targets or restrictions of the natural scientific knowledge. Such is the case, for instance, 'when metaphorical language is employed in a scientific context,' in which case such metaphors 'play a role in the development and articulation of theories in relatively mature sciences. Their function is a sort of catachresis -that is, they are used to introduce theoretical terminology where none previously existed' (R. Boyd, 1979: 357), such as, 'with categories of the mind which could change with time as the accomodation of language and experience proceeded' (T.S. Kuhn, 1979: 418-419). Likewise, the subjects matters of social sciences are mediated (not just by the social discourse of which they are part) by the own history of the particular scientific discourse which asserts, at least in its failures, the absences and inconsistencies in the formalization of some particular social science, and consequently, in the explanation of the social phenomena under consideration. Thus, anticipations of problems, knowledge targets and restrictions (the Bachelardian epistemological obstacle), which are imposed on the researcher as overcoming constraints, are so constituted. 'All this means, of course, that there is no point to the effort to get closer to ultimate truth, mathematical or otherwise. What is needed at each stage is just that one's thinking and communication be in harmony with the whole context, experimental and theoretical, to which it is relevant. And the key to such harmony is sensitivity to disharmony in what has already been done' (D. Bohn, 1974:390). Cf. also #35, #36, #64 to #70 and #73 to #76 and their notes.

(xi/#14) See also #88. Peirce considers research, after asserting its identification with the 'state of doubt', as follows: 'From this conception springs the desire to get a settlement of opinion [that] is some conclusion which shall be independent of all individual limitations, independent of caprice, of tyranny, of accidents of situation...,-a conclusion to which every man who should pursue the same method and push it far enough. The effort to produce such a settlement of opinion is called investigation. Logic is the science wuich teaches whether such efforts are rightly directed or not' (Peirce: 7.316). But cf. #70 and its note.

(xii/#17) The problem usually arises from the accuracy or inaccuracy of the statement that the language of social sciences is the natural language, that is to say that the language used in social sciences is the same as the one in everyday language, with the only distinction between an informal scientific discourse and a formal scientific one, in terms of the features of the 'use' of the natural language (L, Bloomfield, 1973: 99); or if, in spite of the formal analogy between the language used in texts of social sciences and the one used in the social discourse, the former can be classified as really 'natural', as opposed to the language of sciences of nature or of matter, characterized by the use of their own semiotic systems (J.-Cl. Gardin, 1987a: 38; who answers his own question with the assertion that, in the case of social sciences, it is a 'factual opposition' due 'only to the primitive state where most of them are, on the way to an efficient symbolization of the events under study'; thus, it becomes contingent what we consider of a deep epistemological root: the semiotic quality of the phenomena studied in social sciences). The researchers of discourse analysis are the ones who have asserted and assumed the theoretical and operative consequences of the metalinguistic character of the scientific texts they produce, either as an intervention in a 'preconstructed' (that is what is sent back to a previous, external, in any case independent construction, as opposed to what the enunciation actually 'constructs' (P. Henry, quoted and assumed in M. Pˆcheux, 1975: 88-89), or resorting to a 'descriptive metalanguage' inherent 'in texts on discourse analysis as a kind of discourse whose object is the discourse (or discourses)' (J.-M. Marandin, 1977: 55 and 33). Semiotics itself has been considered as a metalanguage both in classical texts: 'Semiotics as a science makes use of special signs to state facts about signs; it is a language to talk about signs' (C. Morris, 1969: 86), and in more recent epistemological contributions; thus, asserting the interpretative character of semiotics, H.Parret considers this one as 'a metadiscourse transposing sense by description', summing up as a work proposal: 'the theory of understanding should be related to the making-to-know of semiotics. In order to traslate these epistemological insights into methodological advances, I want to treat depth and generativity by linking these notions to both construction and reconstruction.It will be the case that semiotic generativity and depth are radically specific with regard to the generativity and depth we construct by translative transposition, on the one hand (in ordinary language use, for instance), and with regard to the generativity and depth we reconstruct by explanatory transposition, on the other (in social sciences, por instance)' (H. Parret, 1983: 78 and 81). From the methodological approach herein and in Parret's terms, semiotics constitutes itself as a set of rules (2nd degree metalanguage) that controls the result of the transformation of the translative transposition task (through which social discourse builds itself: object language) in an explanatory transposition (with which social discourse is reconstructed in the scientific discourse text: 1st degree metalanguage); being the latter task, specific to social sciences as it explains the social effectiveness of the representations/interpretations of such social discourse.

(xiii/#17) L. Hjelmslev, using the logical development of his time (circa 1940; with special reference to the work of Polish logicians, among whom he specially recognizes, in connection to the notion of metalanguage, Alfred Tarski: cf. A. Tarski, 1956) claims that 'one is prepared for the existence of a semiotic [sic] whose content plane is, in itself, a semiotic. This is the so-called metalanguage (or, we should say, metasemiotic), by which is meant a semiotic that treats of a semiotic' (sic) (L. Hjelmslev, 1963: 119-120). However, by following later logical development (cf. specially H. B. Curry and R. Feys 1967: 58-61), so as to avoid the danger of the autonymous way of speaking -the way a token of a symbolical expression is used as the name of that expression-, as well as to make sure the confusion between use and mention has been avoided (Ibid.: 49-50), it has been considered convenient to distinguish between 'metalenguage' and 'outside-language' (a language that is outside some other language) (J. P. Descl‚s, et Guentcheva Descl‚s, 1977: 2) demanding as differential criterion that in the first case the same language, and in the second an artificial symbolical system, and as such 'external to a language', should be used (Ibid.: 5). In this sense, if social sciences develop themselves formalizing (see #58 to #60) the natural language to account for discourses issued in such natural language, they will be built as proper metalanguages; whenever they use a symbolical (artificial) language to account for the discourses issued in a natural language, they will be built as languages that are outside the language under study. The difference is important, as through a metalanguage 'a language can say everything, and particularly describe itself' (Z. S. Harris quoted in Ibid.: 9), while the resource to an outside language responds to the fact that 'for Saumjam, to study the features of a language (or languages) -or U system, object of study- means to build another system, let's call it A, whose aim will be 'to simulate one (or several) language(s)' (Ibid.: 18). This aim of simulation implies the possibility of explaining the procedures through which the natural language (its discourses) under study makes up the meaning of the phenomena it is talking about, whereby the concept of an 'outside-language' is more fruitful than that of 'metalanguage', in spite of their having a logical substratum in common. Instrumentally, this becomes significant when, in the specific investigation, the approach aims at the elaboration of artificial intelligence programmes, that is, at the 'analysis of reasonings proper of human sciences, as for the new perspective opened by expert systems' ( J.-Cl. Gardin et al., 1987b: 1). Due to its importance for the understanding of the global approach of this work, where all the methodological steps tend to make possible the constant critical consideration of the application of reasoning to the progress and the conclusion of any research work, the precision stated by J.-Cl. Gardin becomes of interest: 'Which is the object we name as such when we talk about the study or analysis of reasoning? An easy though incomplete answer is that the concrete expression of some particular reasoning in science is the scientific text itself, where the author states the mental operations which have led him from the observation of certain empirical events to the statement of propositions named in different ways: theses, hypotheses, interpretations, conclusions, comments, explanations, etc.' (Ibid.: 4). However, for communicative easiness, we will herein constantly and exclusively make reference to the metalinguistic features of the 'scientific text' on social sciences, without making use of the difference between metalanguage and outside language. Cf. the note to the following paragraph too.

(xiv/#17) The difference between metalanguage, metatongue, metatext and metadiscourse is not either developed herein as, in spite of their theoretical importance, they would make these notes too confusing and/or obscure the possible use of methodological orientations of the text (one of the most complete developments can be found in the quoted J.P. Descl‚s et Z. Guentcheva Descl‚, 1977); neither do we develop the non-existence of the object language (in its strict sense, because the referents of any language are just semiotic signs or objects and not things, events or phenomena in themselves), being this remark enough to understand the sense with which this expression is used (what results in the fact that even matural sciences use a metalanguage in its textuallity; the only possible object language is that of the phenomena as indexical semiosis; that is to say, as being already signs, but still not-said). However, the expression 'objet language' will be provisionally kept to refer to the social discourse (not to the social sciences' discourse) and to the (non-methodological) discourse of natural sciences. The social sciences' discourse will be thereby considered (1st degree) 'metalanguage', feature shared by the methodological discourse of natural sciences. The methodological discourse of social sciences will be considered (2nd degree) 'metalanguage', feature shared by the epistemological discourse related to natural sciences. Thus we can consider the epistemological discourse realted to social sciences as a (3rd degree) 'metalanguage'.

(xv/#18) 'To the degree that any metalinguistic proposition relating to natural languages is analytic' (J. Lyons, 1977: 292), the character of the (analytical) operations that shape a methodology in social sciences is a metalinguistic one. The reflection upon the metalinguistic features of such methodological operations (which are herein asserted as 2nd degree because of the requirement of having to give simultaneously, whether consecutively or cumulatively, an account of the relations inherent in the social and theoretical discourse) is still to be carried out, at least, according to the information available up to now. There is, however, total awareness of the need to work (metalinguistically) with 'a chain of explicitly defined operations which relate them' (the theses) 'with the events' (J.-Cl. Gardin, 1987b: 6); there is also awareness that 'the development of expert systems, in human sciences as in other fields, is essentially linked with the progress of the so called 'Knowledge Representation and Treatment' techniques (Ibid.: 12) (also of a clear metalinguistic character); therefore the programming constraints guarantee the accuracy of the applied analytical operations, thus not fearing assuming a 'non-philosophical, rather technical, technicist, even technocratic perspective' (Ibid.: 10). Methodological reflection in social sciences, therefore, finds itself forced to modify the domain of its criteria and habits of validation; which will only ground on mere considerations of logical reasonableness, so far as such logical reasonableness is embodied by the possibility of actual applications: 'We should now try to provide that machine with mechanisms which will allow it to control its state of knowledge so that it can conceive reasoning strategies to be carried out through the realization of a given task. The major reasoning operations on reasoning are: -to prove that it can be proved (...); -self-learning (...); -the search for fine solutions (...)' (J. Sallantin, 1986: 31), all of them being operations which imply a practical metalinguistic reflection. But (we must make clear as well) analytical suggestions on the field of artificial intelligence often derive from a world (or from its phenomena) which appears to be a representable one, according to a 'natural logic' or to interrelated 'semantic primitives' (R. Jackendoff, 1987: 91 passim). In this way, it is avoided considering the natural language, as bearer of that logic and provider of that semantics. Thus, the methodological language would be set free from disagreements among natural languages and, of course, from history. The researcher's observation is privileged as a source of safe (therefore supreme) rationality which will also be (guaranteed by the researcher's honesty) universal (or common to any mind) and timeless (or eternal); against this, cf. #28 to #32. This world 'seen by a Martian' (paraphrasing F. Gadet's and M. Pˆcheux's criticism to Chomsky, 1981: 219) or that arbitrary, and therefore dispensable language, 'but only to Sirius's impassive eyes' (as . Benveniste, 1966: 51, ironically says), is in fact much more unreliable than the contingent but positive (empirical) discourse which names it (the world) and uses it (the language). Automatic reasoning on reasoning becomes effective and explanatory when reasoning on the one who reasons is materialized in utterances actually in use in a given society. If this is completed with a theoretical language which is gradually elaborated (formalized) by scientists along the (short or long) history of every particular science, and which provides the concrete reasonings with which he reasons, it will have located the methodological operations in the relationship which makes them efficient in social sciences: the (metalinguistic) meeting point between the theoretical language and the social discourse. Social scientists also have the advantage to count on the rich and long discussed experience that accompanied, in the epistemology of natural sciences, the modifications of the 'rules of correspondence' as 'descriptions of different experimental procedures to connect theories with phenomena' (F. Suppe, 1979: 134); which, mutatis mutandis, is applied to the analytical (metalinguistic) operations which this work discusses.

(xvi/#20) When we differentiate data from information, on the one hand, the distance between this methodological approach and the 'dry-barren land empiricism' which N.R.Hanson refers to, (1977: 21 and 27) is widened and, on the other hand, we fit concepts accurately in terms of what is perceptually observable and what somehow implies some degree of explicit transformation (inference). As regards the first aspect, what P.Suppes asserts is conclusive: 'The doctrine that I want to preach is this. Consider the classical philosophical theses that an absolute causal account can be given of phenomena, that ultimate laws of a deterministic sort can be gleaned from natural phenomena, and that some rockbed of perceptual certainty is necessary to gain a firm knowledge of the world. All three of these theses are false and hopelessly out of date in terms of the kinds of theories now coming to dominate science (...). When it comes to matters of knowledge, real houses are always built on sand and never on rock' (1974: 283). It is advisable to note that the theses expressly denied by P.Suppes are the most frequently used to criticize the absence of scientific precision on social sciences and that Suppes states them at the end of a paper in which he analizes deeply the problem of data corrigibleness, as a critical analysis prior to its identification and use. But the ease to recognize how feeble scientific knowledge is (either natural or social) does not justify the intuitive and non-critical record of the social phenomena that will be studied, thus rejecting their aprioristic character and conditioning them to the features and qualities of the measuring that of them records what is herein called 'information'. M. Borillo, in search of a different scope to the problem of rendering the proper exactness to the analysis of social sciences, suggests 'replacing the information, as it is used in the traditional reasoning of Human Sciences: multiform, irregular, implicit (...), by data, in the strict sense the term has in the Sciences of Nature' (1977: 7). Here data and information would name the same set of entities, the only difference between them being the rigorousness of the reasoning with which they are considered. The change of name aims at provoking a radical change of attitude ('un tel bouleversement') that M. Borillo identifies with the fulfillment of what he calls 'minimum condition of regularity ', in the sense that 'the correspondence between the phenomena and their representing symbolic systems must be such that two identical phenomena must necessarily have the same representation, and that two identical representations must stand for the same two equivalent phenomena (Ibid.). Cf. # 73 to 76. We have herein chosen to keep separate the terms 'information' and 'datum', not by assigning one to social sciences and the other to natural sciences, but limiting the second (datum) to the record of the 'social discourses' (its problem will consist of setting up the proper, necessary and sufficient corpus), and the first (information) to the record of certain features and relations embodied in such social discourses (whose problematics will be concerned with their identification, contrast and transformableness, with the help of the relevant analytical operations; cf. #51 and 64). Cf. the notes to the following paragraphs too.

(xvii/#21) The concept of datum suggested in this paragraph shows a double aspect: on the one hand, its differentiation from the phenomenon and, on the other, its differentiation from the information (for more details on the second aspect, see the notes in the next paragraph). If, in scientific social research, the first task to be carried out by the researcher consists of the compilation of data, and these are not the phenomena, these latter would be excluded from the research; however, what it is aimed at excluding is the trivial, intuitive and holistic consideration of the phenomena, i.e. the concept of the phenomenon as such. For this reason, the compilation of the social device (the social discourse) that reports them is registered as the primal activity of any kind of research (see #24 and the note on it). The phenomenon physicalness, as guarantee of basic objectivity for the investigation, is a fallacy. Objectively, the phenomenon is the way it is represented and interpreted by a certain society, what is been defined somewhere else as 'Semiotic object' (J.A. Magari¤os de Morentin, 1986: 151,152). Such dependence of the objectiveness on the social is one the anticipations stated by C.S.Peirce, and has contributed to its rediscovery: '(...) the very origin of the conception of reality shows that this conception essentially involves the notion of a COMMUNITY, without definite limits, and capable of a definite increase of knowledge' (5.311); '(...) So the social principle is rooted intrinsically in logic' (5.354), or, as he said in another place, 'Logic is rooted in the social principle' (2.654). But such representation and/or interpretation is neither objectively nor straightforwardly cognoscible (see paragraph 33); when it has been attempted to start research with the phenomenon straightforward description, what has been done is to start off with its representation/interpretation the way the researcher perceives it intuitively. As a result, investigations in social sciences have fallen into considerations predominantly subjective and/or literary (see pargraph 62). That is to say, the phenomenon is not herein excluded in order to favour the mental construction of its representation/interpretation, but it is left out just to avoid starting off from a non-critically assumed representation/interpretation. To such purpose, it has been considered in fact objective to take as datum the texts of the social discourses that talk about the phenomenon as an empirically observable physicalness. The total number of these texts acctually available in a certain society contains the total number of possible relations used by such society to build up such phenomenon (see M. Foucault, 1970: 50 ss, for the utterance 'la formation discoursive', akin to this subject). This treatment of discourse as physicalness has been possible by virtue of considering language as physicalness, and not, according to Saussure's classical opinion, as a 'social link', consisting of 'the sum of verbal images stored in every individual' (F. de Saussure, 1972: 97). Such 'sum' is not immediately attainable; what is immediately attanaible is the set of discourses obtained at a certain moment in a given society. That is why . Benveniste has been able to write that language 'functions as a sense yielding machine' (1974: 97). This 'theoretical conversion which takes into account the historical existence of "discursive physicalnesses" leaves out, in turn, the issue concerning the subject and the data, since, beyond the sequence syntactical analysis, it tends to replace the semantic interpretation (...) by the practice of interrogation of the texts with regard to their position in the historical field' (F. Gadet et M. Pˆcheux, 1981: 170). The weakness of social sciences lies in the fact that they start with the 'semantic interpretation' (referring to the discourses or to the phenomena whose interpretation was not questioned, as it was obvious). On the other hand, the possibility of strengthening the scientific structure of the social sciences is founded on the 'practice of interrogation of the texts' (not in a hermeneutical way but in a positivistic one), which, from this point of view, becomes the primary data of any investigation. In other words, if the concept of phenomenon is to be preserved, the phenomena studied by the social sciences are the texts or types of discourse produced by a certain society. They represent, of course, a special kind of phenomena since they are phenomena that build up the signification of other phenomena, having been (and still being) the straightforward access to such signification an illusion in the history of awareness. Discourse analysis carried out by means of computer science has increased the concern for the empirical register of the sequences (phrases) that produce significance and not the other way round. The global view to approach the discursive facts 'has been summed up in the definition of "discursive morphology". The meaning of morphology herein being dealt with, is quasi-naturalistic: the aim is to determine, observe, classify a plurality of "forms" as "groups of features" that would permit to isolate and recognize not "species" but the determining discursive functioning' (A. Lecomte et P. Plante, 1986: 92). Only this reversal in the consideration of what is understood by language and discourse makes it possible to give to social discourse the empirical and objective character of data.

(xviii/#22) The classical concept of information as a dimension capable of being measured by an isomorphous expression of the negative entropy (L. Bertalanffy, 1976: 93; taking such negative entropy or information as a measure of the order or the organization, since such negative entropy as compared to the ramdon distribution, is an improbable state, Ibid.: 42) does not contradict (at least, as it can also characterize) what in this work has been named 'information'. In fact, information (thought as the representations / interpretations identifiable in data) consists of a measurable (computable) order among the physical relationships or distinctive characteristics which can be perceived at the level of language expression (B. Malmberg, 1969: 6), and therefore displayed (even though they require certain analytical operations for their establishment and the description of their operativity) in social discourse. B. Malberg studies information in relation to phonological distinction in as much as the sequence of content has been segmented into a series of entities of discreet content (Ibid.: 106). In this work, information has been considered syntactically and the identification of the corresponding discreet entities will be acceptable or not in as much as they allow or not to establish (be it by identification, contrast or transfer, that is to say, through paraphrastic o subtitution relations) the semantic possibilities of representation / intepretation updated in each realized social disourse; this goes along with the systemic concept of information regarded as 'measure of the organization' in the production of meaning, confronted to 'the ramdon distribution' of the other various relationships that exist in the same discourse but which do not lead to the significance being studied. This exposition is also very much close (and it is applicable, in the methodological practice) to the proposal of a formal grammar of a 'non-ambiguous language' of an essentially syntatic nature, as formulated by R. Montague (1977: 158 ss).

(xix/#22) All the information (identifiable, contrastable and transformable, cf. #77) provided by a certain data corpus (which the researcher considers adequate, necessary and sufficient, cf. #36) makes up a domain, in the sense D. Shapere uses this term; being equally pertinent to the present methodological reflection the conditions required for the association of the information items in such domain: '(1) The association is based on some relationship between the items. (2) There is something problematic about the body so related. (3) That problem is an important one. (4) Science is "ready" to deal with the problem. (...) I called bodies of information satisfying these condictions domains.' (D. Shapere, 1974: 525). In this methodological proposal, the problem consists specifically, in the recognition of such information as producer of the representation/ interpretation with which the social significance of a phenomenon is built up; the building up of that social significance being the importance of such problem.

(xx/#22) When we speak of 'information' in the sense given in this paragraph, the first aspect to bear in mind is what concerns content analysis. As J.-Cl. Gardin says, 'every project of a general method of content analysis in human sciences (...) is a contradiction in terms (...). When Philip Stone and his collaborators chose to call General Inquirer the program of content analysis which would bring them certain fame, and when, in the voluminous book published in 1966 under that title, it was stated that the same computer approach had been used to clear up the content of such different texts as eskimo tales, American political discourse, the last writings of suicidal people, being authentic or not (both men and texts) and many more, a hasty reader could conclude that undoubtedly there existed from now on a 'general method of content analysis in human sciences'. But, 'the method of content analysis is reduced, in practice, to an indexation based on the use of an automatic dictionary or thesaurus, different for every application; therefore, there is nothing general here' (J.-Cl. Gardin, 1987a: 61-62). Gardin himself continues to develop the history of the automatic treatment of information when he recalls that 'the first papers on text analysis (...) established as their own aim the construction of "question-answer systems" capable of detecting certain sense relationships between the statement of questions formulated in a natural language and the data bases made up of texts or fragments of written texts in the same language'; this, notwithstanding the progress it implied as compared to content analysis, made clear 'the inability to account for the simplest sense equivalent concepts among certain statements of any language if a metalanguage that implies -in an almost tautological way- the implicit statement of equivalence, is not defined (Ibid.: 74-75). Finally, the author mentions the coming into light, in 1974, of 'a Minsky report whose title, substance and, above all, the frequency of the quotations it provides, would make it appear as a first manifesto on theoretical positions of text analysis. It is worth remembering at least the title: "A (general) frame for the representation of knowledge". The central problem of text analysis could not be otherwise more exactly and concisely defined. Minsky's objective, however, in his own terms, was less a theory of language than, "a partial theory of knowledge as it is shown through the production and understanding of the natural language". Essential to the thesis is that, in all activity of this kind, the subject necessarily goes to the data in the memory, structured in a certain way. It is still necessary to specify what these data are (semantic component) and the shape in which they are organized (logical component) (...). It is stated here that the kind of semantic logical structures that are moved by the production and understanding of the natural language are finite in number, from the point of view of their form and that it is possible to characterize them, independent of content. Most of the theoreticians of the artificial intelligence applied to text analysis, following Minsky, have enrolled in the last trend' (Ibid.: 75-6). The present work follows also this direction. Thus the concept of information can be understood in this historical and analitically projected context as the set of marks and identifiable syntactic relations in the social discourse (no distinctions are made, for the time being, between 'discourse' and 'text', but the expression is taken in its whole physical sense, so that its empirical perceptual presence must be required of what is colected as 'social discourse' or data; not its reference to an ideal totality) in so much as they are socially used to produce the semantic (representative/interpretative) components of the meaning of the phenomena; and that, therefore, they can allow a data base to be built which will represent, in turn, the expressive potentiality of a society. These marks and relations will identify and interrelate among themselves (by contrast and transformation) through the pertinent analytical operations, according to the conditions sketched in #72 to #89, offering, in this way, an objective base and a process liable to the criticism of the explanatory statements of a certain phenomenon.

(xxi/#22) The relation of information with the representation / interpretation is the main challenge for the fruitful use of the concept of information. The possibilities of success and the risks of failure meet at this point: the essence of information resides in the configuration or organization of signals, independent of their substantive nature (W. Buckley, 1978: 223). Information, in as much as it is a mark or syntactic relation is empirical and observable; in as much as it is producer or carrier of representation / interpretation, it is conceptual and inferable (cfr. #33). The researcher can only choose (and this, after all, is what makes of him a researcher) to build rigorously (according to the constraints of logic and theoretical language of his particular science) the chain of statements going from information (as set of premises) to the representation / interpretation (as conclusion). In the current trends, the problem is worked out in terms of artificial intelligence. This provides the additional control of the processing system: 'one must remember that the processing of an "object", a "problem" or an "event", the moving "of a mechanical arm", a "pawn" on a chess board, the understanding of a sentence, the analysis of a molecule, if produced by a computer, are all said to be prototypical of artificial intelligence if (1) these objects are not processed in themselves (that is, in their basic materiality), but under a symbolic representation of one type, and (2) the manipulation of the elements of this representation (that is to say, the production, the recognition or transformation of these symbolic elements) is controlled in some way by rules specific to the systems used to represent these "objects"' (J.-G. Meunier, 1989: 53-54). Meunier's observation in (1) is synonymous with the difference established in this work between datum and information: datum, as perceptible materiality, it is not processable by itself, but it requires the translation of certain relations it contains to some symbolic representation; that means that information consists of the set of representations / interpretations identifiable in data. On the other hand, agreeing with condition (2), the treatment (identification, contrast, and transformation; cf. #77 to #86) of this information needs certain analytical operations, methodologically basic and specifically suitable (or suitable to the texts of the particular social science the researcher works in).

(xxii/#24) In the wide spectrum of definitions of 'discourse', it is preferred, provisionally, the one that states discourse is the 'textual product of the cognitive operations, socially acquired by a certain sender, in whom, by means of such operations, the meanings of a possible semiotic world for that sender, are built' (J.A. Magari¤os de Morentin, et al., 1989: 22). Such world is wholly coherent with the present methodological development.

(xxiii/#31) As backgound to the present work (which is not stated explicitly anywhere but in this short comment) there exists the decision of transdisciplinary integration among: i/ semiotics (in the widespread sense ascribed to it by Ch.S. Peirce: 'Logic, in its general sense, is, as I believe I have shown, only another name for semiotics (semeiotik‚), the quasinecessary, or formal, doctrine of signs' (2.227); ii/ cognitive science (whose main principles are: '1/ the traditional duality between mind and brain must narrow until it disappears (...); 2/ mental processes can be simulated artificially by man (...); 3/ knowledge is a symbolic representation of what is real'; F. Rastier, 1987a: 6); and iii/ artificial intelligence ('intellectual disturbance' that has been compared 'to the one that accompanied and followed the constitution of Copernican astronomy and Galilean physics', and in which 'a materialism of thought without concessions will finally realize', P. Henry, 1986: 299).

(xxiv/#33) The possibility of having access to the representation / interpretation of a phenomenon, just as it has been produced or registered in a certain social discourse makes it necessary for the researcher to build (manually or through computers, with evident advantages in the second case and in an almost surely insufficient way in the first case) an expert system whose architecture will have to be 'based upon two essential elements: -Knowledge base, which incorporates a subject descriptive representation of known facts and facts that can be deduced. -Inferential engine, a device capable of answering and explaining questions, founded on the structure of knowledge base' (J. Cuena, 1985: 496). The inferential reasoning with which such inferential engine is programmed uses, in general, 'a formulation of Bayesian probabilistic origin, mixed up with aspects of fuzzy logic' (Ibid.: 502; and for 'Bayes' theorem:

P(e/h) . P(h)

P(h/e) = ---------------

P(e)

being (...): P(h/e), the probability that h takes place being supposed the evidence e; P(e/h), the reciprocal; P(h), P(e), previous (a priori) occurrence probabilities of h and e', Ibid.: 510; cf., also 'Moteur d'inf‚rence' in Glossaire, J.-L. Le Moigne (ed), 1986: 359). It is interesting -for it may be puzzling- to observe that H. Parret sustains that inference in discourse-bound rationality does not correspond with logical inference: 'The strategies', in the production of discourse meaning, 'are regularities externalized by a communicative competence: they are chains of reasons and thus based upon processes of reasoning. Discourse, for the pragmatician, is a totality of regularities (recognizable because of theirágenerality) which expresses theoretical and practical reasoning. These strategies of understanding are inferential (not logical inferences, however, because they are realized in and by means of natural language use). Inferential activity is, in fact, a procedure of transposition of meaning from one object level to another periphrastic level of discourse', H. Parret, 1983: 99). It is interesting to notice that the inference Parret is talking about is the one used by the speaker in the production of meaning; that is, the inference the analyst must discover in social discourse, in the sense that it has been placed there by the user of the social rationality; it is one of the pretensions of validity established by the speaker: the pretension that the speech act is correct in relation to the existent normative context (or that the normative context, in whose fulfillment the speech act is performed, is lawful) (J. Habermas, 1989: 144). But of a very different nature is the inference the analyst needs to carry out in order to reach the inference, applying his specific rationality, that through these inferences, effectively used by the speaker, a certain meaning of a certain social phenomenon is produced: 'If, on the contrary, we take seriously the pretensions of the actor in exactly the meaning this rationaliterágives them, we submit his assumed success perspectives to a criticism based on our knowledge and on our comparison of the fatic course of the rational action in accordance with aims to a course built on typical-ideal terms' (Ibid., 166). What Parret refers to is the rational (pragmatic and social) inference established by the speaker; what this methodological sketch refers to is the rational inference established by the researcher, which must fit the constrains of logic and the specific social science whithin whose frame the researcher works. The researcher cannot allow himself to share with the speaker what M. Bunge has called 'catalitical inference' that 'is realized at the level of "anticipation" or "conjecturing" (undoubtedly often incorrect) result of arduous demostrations or demanding empirical demonstrations lacking Ersatz' (M. Bunge, 1965: 125). What has been expressed, leaving aside Bunge's observation on the possibility of 'incorrectness' of catalitical inference, reassures that resorting to logical inference as natural deduction or as probabilistics cannot simply be left out. Even though the scheme of this work does not coincide entirely -as the reader may notice- with the following development by R. Boyd, notwithstanding the fact his presentation has parts in common, thus making easier the understanding of these cognitive operations and of the referent construction: 'In this regard, it is worth remarking that what occurs is not really a division of linguistic labor at all. Instead, what is involved is the social division of mental (or, better yet, cognitive) labor: some of us are auto mechanics and know what "accelerator pump" means, others of us are nurserymen and know what "beech" means, whereas still others are physicists who know what "black hole" means. This division of labor is not primarily a linguistic phenomenon, nor is it primarily an epistemological phenomenon: Instead, as Putnam insists, it represents facts about social organization of labor at a certain stage in the historical development. The division of cognitive labor is related to the issue of reference only because it is reflected in the ways people have of gathering information about features of the world, and because the notion of reference is essentially an epistemic notion' (R. Boyd, 1979: 388-389).

(xxv/#36) As it is proposed in this paragraph, the principle of relevance, without specifying its components, is being taken into account by semiotic analyses, even by those which, following A.J. Greimas school, are still connected to a predominantly intuitive proposal; thus J. Court‚s asserts that semiotics 'faces its objects of study from a well clear-cut point of view common to all of them (...) and that is the principle of relevance: when the object is a data collection, the semiotics will be exercised only in as far as its common characteristics are kept (...). Semiotics postulates that the approach to meaning is only possible by means of different and various approaches, that is to say according to different levels defined by distinctive features common to (or derived from) the objects under consideration' (J. Courtes, 1980: 34). Instead of talking about levels (which imply a cumulative succession) we herein prefer to talk about "information" (to take advantage of the relative neutrality of the terms, with the purpose of issuing a conjunction in simultaneous cumulation of relations or Courtes's 'distinctive features' which will explain the production of meaning, and specially because of the interdependence of such information of the hypotheses stated by the researcher (in a close sense to that established by D. Marr for 'understanding complex information-processing systems', in 1982: 19; cf.also #40 and #41). This is important, as their 'common features' are not so in an abstract orágeneral way, but in terms of the production process of a given meaning whose explanation is pursued. The concept of 'relevance' has a precise status in classical epistemology; P. Achinstein develops it as follows: "Here I must introduce the concept of relevance and speak of a property as relevant for being an X. By this I mean that if an item is known to possess certain properties and lack others, the fact that the item possesses (or lacks) the property in question normally will count, at least to some extent, in favour (or against) concluding that it is an X; and if it is known to possess or lack sufficiently many properties of certain sorts, the fact that the item possesses or lacks the property in question may justifiably be held to settle whether it is an X' (P. Achinstein, 1968: 6), specifying, further on, the qualities of the 'semantic relevance', either with the purpose of classifying X, or of confronting it with other similar substances (Ibid.: 10-1). Note however that in this work, the features of certain data (social discourses) are considered as pertinent to the production/record of certain information. Achinstein's development, when trying to establish the relevance of an X, gets into the analysis of the information for the production of the meaning of a given phenomenon. What happens in data evaluation, herein stated, is that it only informs of its capacity to keep the information that such meaning produces, and that the decision about the relevance such data may have or not can only be reached through the analysis of the production of the meaning in question (what is herein distinguished by naming it as the problem of the 'being in force' of the information kept in such data).

(xxvi/#39) The problem of the 'being in force' is directly connected with the field of 'deontology', term which has moved from naming a science of moral as that which is convenient to the largest number of people (as it was for J. Bentham, creator of this designation), to name the normative sciences. Thus in law, the force of the norm is constituted by the conjunction of validity and efficacy, being the latter, in the Kelsenian view, the condition of validity, 'but not its reason. A norm is not valid because it is efficacious; it is valid if the set of laws to which it belongs is, in general, efficacious' (H. Kelsen, 1958: 49). Of course, there exists a legal concept of prevailing which is purely formal, and depends only on the prerequisite established for the promulgation of a norm. In legal philosophy, the problem is immediately related to that of the existence of Law and so 'the difficulty emerges from the very beginning as a first opposition between positivity and being in force of Law'; if we start defining Law as norm, 'the positivity appears as something added to the being of Law' (C. Cossio, 1954: 181); but for the egological theory (directly derived from Husserl) 'positivity is part of behaviour itself in the sense that it is a fact and it is plainly its existence, considering that the legal object is also according to the way existence occurs in time', and Cossio quotes and agrees with Olivecrona on the following: 'that Law is in force means that some combined human forces support it. The existence of Law is the same as its being in force. To be in force or the existence of Law is thus defined as an effective fact' (Ibid.: 182). Therefore it is not strange that the topic sould have been dealt with, once more, by the factual deontical modalities of logics, 'in utterances such as: "it is compulsory (...)", "it is allowed (...)", "it is forbidden (...)"' to refer to an act (G.E. von Wright, 1970a: 65), what is linked, in this field, to the analysis of prescribing whose life is therefore the duration of the entailment between one rule-authority and one or several rule-subjects. While this connection lasts it is said that the prescription is in force. The existence of a prescription is not the fact that it has happened as such, but the fact that it is in force. (G.E. von Wright, 1970b: 132). The conventional and classical statements about 'prevailing' are overcome (in the Hegelian sense of the dialectical 'aufheben' as 'process of display of the concreteness by mediation of its removed internal determinations'; S. Karsz, 1971: 22) and therefore, restructured in the heart of another problematic, when M. Focault states his proposal of discourse analysis: 'On the one hand the "critical" set which applies the principle of reversal: attempting to establish the ways of exclusion, of binding, of appropriation, I have just referred to; showing how it has been shaped, what needs it answers to, how it has modified and shifted, what constraints have effectively worked and up to what extent they have been changed. On the other hand the "genealogical" set which applies the other three principles: how the discursive series have been shaped, going across, in spite of or with the support of those systems of constraints; which was the specific rule of each one, and which were their conditions of appearance, growing up and variation' (M. Foucault, 1971: 62. Such is the 'being in force' of the information (as social ways of representation/interpretation of phenomena) which the researcher looks for and recovers in the social discourses, assumed herein in their full positiveness or effective existence and concrete history, within a certain society; leaving out as far as possible and with the greatest care, the idealistic remnants and the apriorities of what seems to be obvious in knowledge, substantial in the essence of phenomena, and universal in reason.

(xxvii/#42) Herein, the expression 'provability of an explanation' does not refer to a probabilistic reflexion on the validity of a given explanation, but to its use in the sense the similar expression 'provability of a theorem' has in logic, concerning its formal deducibility in the predicate calculus (W. Hatcher, 1968: 30). It is therefore asserted that it is the proposal of explanation (of the meaning of a certain phenomenon), which constitutes the first (theoretical) hypothesis of research work, that must be proved by a sequence of statements which account for the productivity of the representations / interpretations identified by the researcher in the corpus of selected social discourses. It is also the sense conferred by J.-Cl. Gardin to 'the nature of operations which support the steps from the early data to the final statements of the construction, where the author's conclusions are stated' (J.-Cl. Gardin, 1987a: 32).

(xxviii/#46) The three possibilities fit the Popperian criterion of falsifiability (different from the Peircean fallibility, to which we will refer in the note to #70) as conditions to decide the empirical character of a system of statements, thus making of such a system a scientific one (always within the Popperian assumption) because it is possible to refute it through experience (K.P. Popper, 1977: 40, 75, 108, passim). It has to be proved whether each of the statements included in the expressions of each hypothesis turn out to be false or not. That is, at the end of the investigation there will be new statements that will be expressed (in connection with the corresponding hypothesis) according to some of the epistemic modalities de dicto, taking into account that 'if the negation of one proposition is verified, the proposition will be called falsified' (G.H von Wright, 1970: 55ss). Of course they are different sorts of falseness: in the first case (falseness of the 3rd hypothesis: work hypothesis) what is denied is that the compiled social discourse possesses enough information to prove the explanation of the social meaning of the phenomenon (therefore it is a case of empirical falseness); in the second case (falseness of the 2nd hypothesis, the methodological one) what is denied is that the information (the relation marks making up the representations / intepretations) identified in the compiled social discourse may match the reresentations / interpretations with which a given society builds up the meaning of a given phenomenon at a given moment (then, we verify that there is a pragmatic contradiction between the statements recovered as information and the statements that actually prevail in the society under study); in the third case (falseness of the 1st hipothesis, the theoretical one) what is denied is that we can deduce the hypothetical statement of the meaning ascribed to a certain phenomenon from the set of factual statements concerning the representations / interpretations (the information) that prevail at a given moment and that have been established through the analytical operations applied to a relevant corpus of social discourses (therefore it is a case of semantic falseness).

(xxix/#47) This paragraph highlights the logical contradiction that must exist between the hypothesized meaning and the social representations / interpretations that are effectively in force in the social discourse relevant to reject the former. In fact, to be able to deny that hypothesis it was necessary to prove the force of the identified representations / interpretations and the relevance of the social discourse that produces / records them; what is denied is that the said relevance and force (both properly verified) are the logical antecedents of the meaning stated in the hypothesis; only in this case will the falseness of the theoretical hypothesis have been proved (it will have been falsified). While the validity of the 2nd and 3rd hypotheses (jointly) is not successfully shown, the assertion or denial of the 1st is not possible (this will respectively happen according to whether the logical derivability of the latter -1st- is or is not proved through the previous two), or whether the logical contradiction between the 1st hypothesis and the other two is or is not proved. The test of logical derivability and non-contradiction as well as the test of logical non-derivability and contradiction are not symmetrical, since in one case, that of contradiction or non-contradiction, this comes from pure reasons of propositional logic (D. Hilbert and W. Ackerman, 1962: 21), while in the other, that of the derivability or non-derivability, it belongs to the propositions of the system (J. LadriŠre, 1969: 153).

(xxx/#47) It is worth wondering, then, whether the methodology establishes an order to carry out research or if this order is the one that corresponds to the document where the results of such investigation are spread out. In more academic terms, the problem has been stated as 'the distinction between context of discovery and context of justification'. P.K. Feyerabend summarizes the classical view saying that the discovery may be irrational and does not need to follow any acknowledged method. On the other hand, the justification , or to use the Sacred Word of a different school, criticism, begins only after the discoveries have been made and follows an orderly way. To this, he contrasts his own answer: Research is, at its best, an interaction between expressed explicitly new theories and old forms of things which have filtered into the language of observation (P.K. Feyerabend, 1974: 99-100). Cf. #56 and #57. This interactive quality of research has been asserted by Ch.S. Peirce when he objected to the existence of intuitive knowledge, and therefore, a priori, as basic epistemologic grounds, what is known as the problem of grounds. The alternative paradigm of investigation and knowledge that Peirce starts to develop in these writings (specially in Questions Concerning Certain Faculties Claimed for Man) and which he specifies and modifies throughout his career, is a conception of investigation as a self-correcting process that has no absolute starting or finishing point and where any assertion is subject to subsequent rational criticisms, though we may not be able to question all the assertion at the same time. Our cognoscitive axioms are not legitimated by their origins, since the origins of knowledge are varied and fallible, but by the norms and rules of investigation itself. And even these norms, rules and patterns are subject to rational criticism (Richard J. Bernstein, 1979: 183). In this paper, the interaction between praxis of research and the congnitive conjectures that give rise to such praxis are conceived in a dialectical way, in what concerns the resolution of contradictions. By means of the overcoming of such contradictions research unfolds the complex frame of knowledge which is gradually producing (Cf. #77 onwards and, specially, #87 and #88).

(xxxi/#51) Such a set of technical operations constitutes what can be called 'rational thinking'.

(xxxii/#52) 'In everyday life we usually use the expression "rational thinking". It is a relatively comprehensible utterance for whose extra-scientific use, it is enough with the representations we have on its meaning. A slightly deeper understanding of what is rational thought results in a non-immediate way from the formulation of an intelligible expression. This understanding is achieved through scientific analysis, that is, by means of the application of a particular way of rational thought to the awareness of an object that, in this case, is precisely the rational thought. Reason emerges and develops as a human social property in inextricable relation to work on one hand, and to language on the other hand, and it has a biosomatic bearer: it lives attached to the brain and to the nervous system of the man who lives in a society" (J. Zeleny, 1982: 11-12). Everything understood in a pluralistic sense of reason which, nowadays, exceeds even 'the concept of reason that Habermas projects in terms of the theory of consensus', excess which has been pointed out by Lyotard: "to recognize the autonomy and specificity of the plurality and intranslatability of language games interwoven among themselves, without trying to reduce the ones to the others; with a rule which would be, nevertheless, a general rule, "let us play (...) and let us play in peace"' (quoted by A. Wellmer, 1988: 109). Something simply and humbly anticipated by the founder of the modern semantics, M. Breal: 'being language people's work, it is necessary, to understand it, to give up the logician and become people with it' (1924: 233).

(xxxiii/#56) Intuition is a social and historical knowledge, very close to 'common sense' and, like common sense, confused, as G. Moore thought, the problem of whether we understand its meaning (he refers to the sentence 'the Earth has existed for many years'), with the problem of whether we know what it means, in the sense of being able to analize its meaning correctly (G. Moore, 1974: 257). This is the kind of intuition the researcher has no right to, since it is, as Nagel and Neuman said, an elastic faculty; our children will probably find it quite easy to accept as intuitively obvious the paradoxes of relativity, as we do not back down before ideas which were considered absolutely non-intuitive a couple of generations ago. Furthermore, as we all know, intuition is not safe guidance: it cannot be used appropriately as a criterion of truth nor of fruitfulness in scentific explorations (E. Nagel and J.R. Neuman, 1970: 29). Of course, it is out of this exclusion the concept of 'mathematical intuition' , by virtue of which Hilbert 'against what Dedekind, Frege, Russell and logicians in general accept (for whom mathematics is a part of logic), asserts the independence of mathematics from logic. Logical operations affect objects that must previously be judged by intuition and which precisely constitute the mathematical reality; J. LadriŠre, 1969: 26), as well as intuitionism which has been Brouwer's answer to the crisis of the mathematical grounds, though in full controversy with Hilbert (H.B. Curry and R. Feys, 1967: 57-58). Therefore, it is necessary to distinguish between intuition as assumption in the origin of knowledge (problem which is gnoseologically valid, whatever the attitude assumed may be) and intuition as methodological resort for the development of research, which is the position rejected herein. But, see note to the following paragraph.

(xxxiv/#57) Intuition as the researcher's empirical component of thought is not rejected either, coinciding in a way with M. Bunge's reflexions: 'In any scientific task, there intervene from the search and issue of the problem to the control of the resolution, and from the invention of the guide-hypotheses to their deductive elaboration, the perception of things, events and signs; the visual representation or imagination; the formation of concepts at various levels of abstraction; the comparison which leads to establish analogies and the inductive generalization together with the crazy conjecture; the formal as well as the informal deduction; refined and rough analyses and probably many other ways of shaping, combining, and rejecting ideas, because, let's mention incidentally, science is made of ideas and not of events (...). What follows is an enumeration of the most frequently accepted uses of the term "intuition" in the contemporaneous scientific literature: quick perception, imagination, abbreviated reasoning and common sense' (M. Bunge, 1965: 88-89). There is probably total agreement with all that; but provided it is proved; what we are claiming herein is that intuition is neither a provatory process nor a provatory resort. Cf. #67 and its note on the Peircean 'abduction'.

(xxxv/#58) It is advisable to distinguish between 'formalization' and 'symbolization'. 'Although symbolization and formalization are two different and theoretically detachable steps, they are in fact closely associated, since the latter is greatly made easier by the former, so that it implies it almost unavoidably' (R. Blanch‚, 1965: 44). When the theory is axiomatized, 'it gives us first sentences which claim, in a symbolical language, logical relations among first terms: as it does not propose them but as hypotheses, we accept them as such, subject to their compatibility. But from there onwards, we will only receive a new term if it is defined with the aid of the first terms; we will only accept a new propositon if it is proved with the aid of the first sentences' (Ibid.: 45); such is the task and constraint of formalization. As regards symbolization, 'the aim when we locate a deductive theory in an axiomatic form is to separate it from the concrete and intuitive significances on which it was first constructed, in order to make the abstract logical scheme appear clearly (...). Thus, it is soon felt the need to subtitute symbols lacking previous sense, and consequently capable of receiving the one that axioms convey them exactly and exclusively, for the words which named the early notions of the theory, even with the weight of their intuitive significance' (Ibid.: 43).

(xxxvi/#58) E.W. Beth's task of differentation between 'formal derivability' and 'semantical implication' becomes important, specially in the field of social sciences; in the first case, there exist certain formal rules of inference, each of which, if applied to the proper premises, produces an immediate conclusion (E.W. Beth, 1978: 6); and, on the other hand, the notion of logical consequence, relevant to this context, can be defined as semantical implication, being depedence on the context, in this case, fundamental (e.g., the context of the set of entities, with their coresponding meaning, selected to interpret the logical variables), what causes that, in the case of the semantical implication, truth-value (that is to say, the truth or falsity) of the new premises and the new conclusion should play an essential role (Ibid.: 7).

(xxxvii/#58) M. Borillo wonders 'Why trying to move towards the formalization of Human Sciences?' (1977: 24) (we will mantain here the most usual French designation for 'social sciences'). The answer, developed through a set of observations suggested by his practical experience and by the investigations carried out for some years in the Laboratoire d'Informatique pour les Sciences de l'Homme, which depends on the C.N.R.S., may be summarized as follows: 1) 'to transfer to a formal expression all or part of the constituting elements of a problem is a constraint which implies more exactness in the analysis of the empirical phenomena and in the ratiocinalistic driving (...)'; 2) 'formalization allows the construction of "more powerful" theories (...)'; 3) 'the problems of "theoretical adjustement" of mathematical methods to the deep nature of the facts under consideration (...) are crucial for a reflection on the scientific status of Human Sciences (...)'; 4) 'the functional aptitude of the formal methods to understand, represent and order information in its extension and complexity allows us to incorporate pieces of information inherent in different phenomena, but capable of finding in their association a richer meaning (...)'; 5) 'it is likely that the changes of problematics, from the point of view of the extension and richness of their contents, is related not only to the functional aspects of the said formal methods, but also and directly to the theoretical nature of the constructions (...). This capacity of integration (..) constitutes the safest internal substratum of the forthcoming transdisciplinarity (...)'; 6) 'finally, the impact of the formal methods on Human Sciences is manifest in the praxis of research and even more so by the fact that the changes which have affected Natural Sciences at a secular scale are about to find their equivalent at decade scale for Human Sciences (...). It is enough to remember the most important points: (...) from the technical point of view, problems are those of the operative enforcement of methods and of the theoretical approaches already mentioned (...); from the social and institutional point of view, it is a fact that scientific production quickly acquires a collective dimension opposed to the individual aspect, to the wiseman's erudite autonomy which has prevailed up to now (...); interdisciplinarity has a double component, thematic (..) but also methodologic, as the formalization will require the work of the semiotician as well as the mathematician, the logician and the scientist computer (..), without neglecting the dialectics of the individual and the collective in the act of reflection and creation (...), the whole group appears more and more as collective responsible for their work. However, experience shows that the diversity of individual roles, far from be blurred, is enphasized by the multiplicity of knowledge and development confronted (...). Another illusion will be to think that informatics techniques may establish by themselves the researcher's free access to scientific information (...). It sould also be mentioned that the embodiment of formalization in Human Sciences will be "difficult", as it may be the change from rhetoric to reasoning, though the nature of this transformation will lead researchers of Natural Sciences and researchers of Human Sciences together to the search of an epistemologic convergence inside the sphere of scientific knowlege' (Ibid.: 25-30). For the fulfillment of these possibilities and expectations (that in some fields of study of Social Sciences are already a reality) several analytical techniques called 'of apprximate reasoning' have converged. These techniques, instead of limiting themselves to 'represent the level of certainty by means of a unique value' propose procedures which aim at 'improving the quality of representation', in this case, of social phenomena, through the acceptance of a 'larger number of parameters to evaluate the level of deductibility of a rule' (J. Cuena, 1985: 523-524). The so-called 'non-monotonous logics', the fuzzy logic, the possibilistical method and the present developments of the mathematical representation of the possible worlds, converge to that.

(xxxviii/#59) Leaving out the complex problematics of scientific definition, it is important, specially for research in Social Sciences to notice the difference analysed by S.A. Kripke, between the scientific definition as synonymous and the scientific definition as fixing a reference: 'Imagine a situation in which human beings were blind or their eyes didn't work. They were unaffected by light. Would that have been a situation in which light did not exist? It seems to me that it would not... the light might have been around; but it would not have been able to affect people's eyes in the proper way. So it seems to me that such a situation would be a situation in which there was light, but people could not see it. So, though we may identify light by the characteristic visual impressions it produces in us, this seems to be a good example of fixing a reference' (S.A. Kripke, 1980: 129-130; emphasis added) for light. 'Perhaps we can imagine that, by some miracle, sound waves somehow enabled some creatures to see... Would we say that in such a possible world, it was sound which was light, that these wave motions in the air were light? It seems to me that... we should describe the situation differently. It would be a situation in which certain creatures... were sensitive not to light but to sound waves, sensitive to them in exactly the same way that we are sensitive to light. If this is so,... when we talk about other possible worlds we are talking about this phenomenon in the world, and not using "light" as a phrase synonymous with "whateverágives us the visual impression-whatever helps us to see"'; for there might have been light and it not helped us to see; and even something else might have helped us to see. The way we identified light fixed a reference' (Ibid.: 130-131).

(xxxix/#62) 'In order to specify the nature of poetry we have to consider now the specific semantic character -being a polisense (polisemic) discourse orágenre- that distinguishes it from science, univocal discourse orágenre' (G. della Volpe, 1966: 120). G. della Volpe so differentiates poetical discourse (in which each term, subject to the action of the surrounding terms of its context, acquires a semantic plus which opens a new sense for it) from scientific discourse; and he also differenciates scientific discourse from ordinary discourse using the terms 'univocal (Cf. Galilean "exact determination") and equivocal (Cf. Galilean "equivocalness") as the cases may be; that is, according to whether it is understood the word or discourse as scientific, or as ordinary and vulgar, using also for the latter the more rigorous term: literal-material' (Ibid.: 121); as it had previously been the case with scientific discourse (in which each term, subject to the action of the other terms of its context, is affected by a semantic minus that assigns a unique and unmodifiable sense to it: the preservation of its explicit definition through its use in contexts which restrict it to such definition). It is particularly interesting that the poetic plus and the scientific minus (which della Volpe characterizes as 'the respective semantic locus' that generates the poetic and the scientific; ibid.: 122) are accomplished as the possible transformations of that literal-material value which ordinary discourse has, as being social raw material of any discourse. That is why a poet does not use his terms and expressions as concepts are used in Social Sciences: 'in a contradicting way' (mimesis, difference, feature, gramma, deconstruction: pertaining to Derrida's theoretical language) 'the equivocal significant azur has many semantic functions in Mallarm‚'s poem, but it cannot be read as a concept. Nobody would dare criticize Mallarme's inexact use of the word azur; on the contrary there is criticism to Derrida for his semiotic concepts and to Adorno for his ideological ones, for being vague though full of sense, when they are used as pertinent to Social Sciences, because their dissolution in 'poetic' expressions causes the disintegrations of Social Sciences' (P.V. Zima, 1989: 12).

(xl/#66) Cf. the already classical controversy whose main protagonists were K.R. Popper and T.W. Adorno (T.W. Adorno et al., 1973).

(xli/#67) The Peircean concept of 'abduction' (which sometimes is called also 'retroduction') offers interesting possibilities to reflect on the raising of hypothesis. As I have in effect observed elsewhere: 'I can attempt to suggest that the charateristics Peirce assigns to abduction correspond to the constraints of rationality inherent when working out a research plan. The fact that he has assigned to abduction the same level as induction and deduction, as one of the "three fundamentally different kinds of reasoning" in science" (1.65), connecting them respectively with "the conceptions of deductive necessity, of inductive probability,of abductive expectability" (5.194; emphasis added), may confuse, but pointing out that "the abductive inference shades into perceptual judgement without any sharp line of demarcation between them (...). The abductive suggestion comes to us as a flash. It is an act of insight, although of extremely fallible insight" (5.181). Nevertheless, abduction would seem to possess an identifiable, definable, and usable internal logical structure as it can be asserted about induction and deduction. In this sense, Peirce points out that "abduction, although it is very little hampered by logical rules, nevertheless is logical inference, asserting its conclusion only problematically or conjecturally, it is true, but nevertheless having a perfectly definite logical form" (5.188). Peirce succeeds, with particular efficacy, in explaining abduction when he relates abduction with the analysis of the "rational conjectures" (1.608), associating it to "the third kind of reasoning [that] tries what il lume naturale, which lit the footsteps of Galileo, can do" (1.630), and, particularly, when he states abduction as a necessary "... sort of inference (...) called making an hypothesis. It is the inference of a case from a rule and a result" (2.623), or "adopting a hypothesis as being suggested by the facts" (7.202). The conceptual device for the working out of a research plan is thus constituted: "it became axiomatical that a hypothesis adopted by abduction could only be adopted on probation, and must be tested" (7.202). Therefore, once the efficacy on the making of hypothesis has been fulfilled, "retroduction does not afford security. The hypothesis must be tested. This testing, to be logically valid, must honestly start, not as Retroduction starts, with scrutiny of the phenomena, but with examination of the hypothesis, and a muster of all sorts of conditional experiential consequences which would follow from its truth. This constitutes the Second Stage of Inquiry. For its characteristic form of reasoning our language has, for two centuries, been happily provided with the name Deduction" (6.470). Thus, retroduction takes part both in the statement of hypotheses from the facts the researcher already handles, and in the preparation of the process of their testing through the observational and/or experimental steps which would require the consequences derived from such hyotheses; and nothing but this constitutes the process of working out a research plan' (Magari¤os de Morentin, J.A., forthcoming: 46-49).

(xlii/#68) The heuristic efficacy of analitical operations is to be understood according to H. Bremmernans's definition, which 'in spite of its apparent negative condition, seems the most acurate today: "Heuristic is any method or programmable guide-principle that allows the elimination of not very promising possibilities in a process of investigation'" (quoted from J.-L. Le Moigne, 1986: 43).

(xliii/#68) The concept of science as construction, in particular with reference to Social Sciences, has been developed by J.-Cl. Gardin and fruitfully applied to his research in the field of archeology: '(...) we must consider scientific texts as objects constructed with particular aims and by means of particular methods' (J.-Cl. Gardin, 1987b: 4). This textual characteristic leads to the notion of symbolic construction, about which he points out 'the two necessary components: (a) on one hand, the semiotic foundation, that is to say, the set of symbols which constitute the representation of the objects under study ("base data", "data base"); (b) on the other hand, the informatical framework, understood in a real or figurative way, that is to say, the total number of operations which links the bottom with the top of the construction (hypotheses, conclusions), realized by means of the algorithmical chains capable of computational treatment' (Ibid.: 15). '(..) the analysis of constructions sets forth only two precise questions (...): if we deal with a certain written work, conceived as the product of a scientific task, we will at times wonder what this written work is made of and what this written work does'. As regards the first subject, the pertinent question is: 'by what means does the author move from a set of given observations to its interpretation, called, according to each case, explanation, taxonomy, rule, law, theory, model or any other term which may imply going through the space that lies between the ingenous apprehension of phenomena and the formulation of a scientific comment?' (J.-Cl. Gardin, 1987a: 29). As regards the second item: 'what a scientific text does', Gardin recognizes some lack of balance between the possible efficacy of 'a scientific construction in fields such as physics, astronomy or biology', in whose fields it grants 'the one who understands it or profits from it -the two uses are not necessarily linked- some power over the kind of phenomena that construction refers to; power in the most concrete sense of the term, as it finishes in a nuclear explosion, the sending of man to the moon or the manufacture of a beneficial bacterium', and the results of social sciences concerning which he asks himself 'which increase of power do we have the benefit of, after having read a scientific explanation of les Chats of Baudelaire?' (Ibid.: 29-30). We may rightly think that Gardin has been driven by habitual constraints in his consideratios of the social science, something that is particularly strange in him considering he is so much interested in the use of expert systems. Scientific knowledge of Baudelaire's poem offers, at least, important information about the operation process of the human mind (e.g., in the creation of language or of new possibilities of world representation) that is the raw material for informatical reflection (e.g., in what concerns artificial intelligence), essential to the programming (even for hardware) which makes possible man's trek to the stars, foreseeing famine in nations and continents, or designing politics to knock down walls, to preserve idiosyncrasies, and to prevent nuclear explosions over populations. The power of Social Sciences is not small, although its consolidation and increase has been (as it has been known since there is written history) the object of hatred of all authoritarianisms. This comment does not restrict the validity of Gardin's theoretical proposal which summarizes with both questions (what it is made of and which is the efficcacy of scientific constructions) and with the only concept of 'scientific construction', what he names 'the field of practical epistemology'.

(xliv/#70) It is already an assumption of epistemology that, no matter what demands may be imposed on the analytical treatment of certain information, there is no total guarantee of the validity of acquired knowledge: '(...) although the scientific development is often reasonable, there is no question here of anything that could appropiately be called a "logic of discovery," for there is no guarantee that a certain line, or even any reasonable line, of research will eventuate in a solution of the problem. Rather than speaking of a logic of discovery, it is less misleading, and more faithful to the spirit of science, to describe the analyses given here as having been concerned with the rationale of scientific development.' (D. Shapere, 1979: 618). But this does not prevent the scientific treatment of the information from applying the greatest possible constraints of rationality. The results of criticism (and semiotics is a particularly priviledged subject in this action) to the logocentrism inherent in the task of formation of theoretical concepts have been positive. Adorno's negative dialectics and Derrida's deconstructive reflection have been specially sharp tools for the discovery of the snares of language in the making of theories, although the greatest requisites of logics and rationality with universality pretensions have been kept . And that universality was the mistake; as it was in Adorno's and Derrida's cases who did not notice that their own criticism also needed to create concepts in order to criticize concepts, what made possible to review their criticism towards a proposal of a 'new logocentrism' (P.V. Zima, 1989: 12); but it justifies even to a lesser degree P.V. Zima's claim to reject such logocentrism resorting to a dialogical thesis in which, according to the specific approach assumed by this author, beneath a contrastive and historical appearance, lies the old essentialism and universalism, herein excluded: 'without these collective universals, the terminology of social sciences would be not only contradictory and fragmentary but also so incommensurable that every discussion among sociologist, semioticians and psychologists would be like a dialogue among the deaf' (Ibid.: 17).

(xlv/#70) One of the criteria that Peirce adopts to reflect upon the main features of knowledge production by research consists in the distinction between doubt and belief. 'When we believe, there is a proposition which according to some rule determines our actions, so that our belief being known, the way in which we shall behave may be surely deduced, but in the case of doubt we have such a proposition more or less distinctly in our minds but do not act from it' (7.313). From that difference (whose conclusive opposition he avoids, however: 'Belief and doubt may be conceived to be distinguished only in degree'; 7.314), Peirce asserts the attitude that the researcher must adopt: 'Living doubt is the life of investigation. When doubt is set at rest inquiry must stop' (7.315). Therefore investigation does not aim to offer conclusive answers, but "it tends to unsettle opinions at first, to change them and to confirm a certain opinion which depends only on the nature of investigation itself' (7.317). The acceptance of the result of investigation as final changes knowledge into belief and leaves out reason: 'The only justification for reasoning is that it settles doubts, and when doubt finally ceases, no matter how, the end of reasoning is attained' (7.324). 'Now if belief is fixed, no matter how, doubt has as a matter of fact ceased, and there is no motive, rational or other, for reasoning any more' (7.325). Fallibilism is a concept which was also specially developed by Peirce '...there are three things to which we can never hope to attain by reasoning, namely, absolute certainty, absolute exactitude, absolute universality' (1.141). '... On the whole, then, we cannot in any way reach perfect certitude or exactitude. We never can be absolute sure of anything, nor can we with any probability ascertain the exact value of any measure orágeneral ratio. This is my conclusion, after many years study of the logic of science; and it is the conclusion which others, of very different cast of mind, have come to, likewise' (1.147). The whole of this 5th point ('Fallibilism, continuity, and evolution', three key ideas in Peirce's thought) of the 3rd chapter ('Notes on Scientific Phylosophy'; in both cases, titles stated by compilers) is of great interest. The subject reappears in 'The fallibility of reasoning and the feeling of rationality' (2.151-174). Nowadays, there is a widespread tendency towards moderation against too much confidence in reasoning and a 'distrust of very strong systematizations and of excessive generalizations, and in general, of normative-content constructions; on the other hand, it is encouraged a positive fallibilist distrust and a tolerance towards the procedures of weak science which do not fit the positive scientific models' (J. Habermas, 1988: 105), all that linked, almost in a contradictory way, with the criticism to 'the rational legacy of the Critical Theory (...) that is to say, with a philosophical concept of truth, taken from Hegel, which the old Frankfurtians have never resigned, and which is unreconcilable with the fallibility of scientific work' (Ibid.: 142).

(xlvi/#70) The concept of plausibility is connected, in the methodologic constraints related to the development of artificial intelligence, with the concept of interpretation. R. Rastier asserts, against the criterion 'of removing the interpretations considered inappropriate, and postulating the univocity of the selected text', (adding the following footnote: 'this rationalistic prejudice goes back to the eleatic philosophy'), 'we prefer another approach. For an interpretative semantics, equivocity is an essential datum. In general, multiple interpretations are given. At best, it may be asserted that one interpretation should be preferred to the rest. In other words, and in spite of the fact that all hermeneutic tradition clashes with such conclusion, the sense of a text does not belong to the order of truth, but to the one of the plausible. Instead of rejecting what we thought inappropriate interpretations, it is convenient to arrange them hierarchically, grading their plausibility in relation to a given strategy' (R. Rastier, 1987b: 100). Plausibility has already a distinct (or plausible?) place among the modalities of the modal logics: 'certainty (assertable or deniable)', entailed to completeness; 'plausibility (justifiable, refutable)', entailed to contingency ('it is not refutable that it is not'); and 'propensity (rectifiable, silence)', connected with monotony, the latter being 'an interesting sort of modality suggested by S. Watanabe in his propensity theory' (J. Sallantin, 1986: 127). Cf. the text by J.-Cl. Gardin, herein quoted several times (1987a).

(xlvii/#77) We herein state as assumption (the foundation of which has been developed in another place: J.A. Magari¤os de Morentin, 1986: 141-59) the assertion that these three operations are necessary and sufficient to produce the explanation of the social meaning of any phenomenon; their recurrent application requires the inclusion of the result attained by each of them in each one of its successors; we do not intend to give a definite answer to the problem of determining when the primitive character is assignable to syntax or to semantics, but to give an operative solution through their mutual integration into a historical perspective.

(xlviii/#78) The identifying operation is directly connected with perception, either in the text or in the phenomenon (through another text). The search for effective syntactic relations (whatever may be the semiotic raw material of the text under study: verbal, graphic, objectual, etc.) relations which, it is hypothetically asserted, are able to produce (with the contingent permanence the history of a given community determines) the meaning that identifies a certain phenomenon, aims to realize effectively some of H. Parret's theoretical concepts: 'Going through surfaces and manifestations means to search for permanence and isomorphism. Fortunately, neo-Hjelmslevian semioticians admit that the search for identity itself has to be grasped as a procedure of identification or of the recognition of identity' (H. Parret, 1983: 86). At this stage, the researcher applies the identifying operation, without considering temporal relations (acrony), but only topological relations (syntax). What is identified, as its result, is the semiotic object: piece of text that efficiently represents/interprets a given phenomenon feature; or a phenomenon feature constructed (or capable of being stated) by a given piece of text (intensional consideration of semiotics): 'Identity (...) is not the relation between an object and itself, but is the relation which holds between two names when they designate the same object' (S.A. Kripke, 1980: 107); identity , therefore, in the designation of the object or in the designation of a feature of that object. With a different terminology, we defined this operation (in J.A. Magari¤os de Morentin, 1986: 145; called there 'attribution') as: 'That one through which a form is related to a value'.

(xlix/#81) This contrastive operation aims to establish the presence of, at least, 'two antagonistic values assigned to the same place of a given syntactic scheme'; presence that results from the analysis of a given corpus of social discourses, so that 'it may originate, in the intradiscourse of the discursive sequences governed by a certain discursive formation, a contrastive modality of identification syntactically carried out" (J.-J. Courtine, 1981: 94). This contrast 'produces a reference effect, in the sense that its syntactic form causes the contrastive identification of synonymic substitutes which belong to two antonymic referential sorts" (Ibid.: 102), being in turn, by this contrasting referentiality, how a phenomenon acquires the semiotic object character, not only identified, but semantified through the effective conjuction of the simultaneously competitive statements (in synchrony) prevailing in a community. In J.A. Magari¤os de Morentin, 1986: 148, this operation is defined (under the label of 'substitution operation') as 'that one through which, given n different universes of forms (n ò 2) in synchronical relation, the forms belonging to one of such universes are related to the values attributed to the forms belonging to other universes'.

(l/#84) The trasformational operation (term which bears, in this case, no relation to the generative of transformational grammars) aims to show the historical value of the social meaning of a given phenomenon, linking the contrast of meanings that prevail at a given moment in a certain society, with the contrast of the meanings that obtain at anotherágiven moment (prior to or subsequent to the other) in that same society. The appearance of such historical value will depend on whether or not the relation between those contrasts can show the effective transformation of the meaning in the two moments under consideration (diachrony). Such effective transformation of the historical value is called 'overcoming' and with that label, it has been used in J.A. Magari¤os de Morentin, 1986: 152, that here we reproduce, notwithstanding its exaggerated length, but giving preference to its operative and descriptive value: 'That one through which, given n pairs of universes of forms (n ò 2) with a structure of systems and discourses and arranged according to an ordered or orderable series, every form belonging to any of the components of a (provisionally) last pair of universes of forms, requires the existence of another form, in any of the preceding universes of forms, affected, simultaneously, by the denial of its own values and the affirmation of the substituted ones; and every value belonging to any of the components of a (provisionally) last pair of universes of values, requires the existence of another value, affected, simultaneously, by the denial of the relation among the forms that attribute such value to themselves and by the affirmation of the relation among other forms identified through a different attribution'.

(li/#87) Both a dialectical consideration and another one that makes use of the topological metaphor of the catastrophe converge in this definition of 'overcoming'. In the first sense, 'what turns out to be overcome is abolished, suppressed in a sense. Nevertheless, in another sense, what was overcome does not stop existing, it is not reduced to the pure and simple nothingness; on the contrary, what was overcome raised to a higher level. Because it has worked as a stepping stone, as mediation to attain the higher "outcome"; and certainly that mediation stage does no longer exist in itself, in isolation, as it was before, but it persists through its denial, in its outcome' (H. Lefebvre, 1984: 267). A scholar so little suspected of hegelianism as W. Labov states the following reflections which help to understand the concept of 'overcoming', and even to establish, from a pragmatic perspective, its empirical usefulness; 'Those variables which are closer to surface structure frequently are the focus of social affect. In fact, social values are attributed to linguistic rules only when there is variation. Speakers do not readily accept the fact that two different expressions actually "mean the same" and there is a strong tendency to attribute different meanings to them. If a certain group of speakers uses a particular variant, then the social value attributed to that group will be transferred to that linguistic variant. Sturtevant (1947) has proposed a general model of linguistic change showing the opposition of two forms, each of them favored by a particular social group. When the issue is resolved, and one form becomes universal, the social value attached to it disappears' (W. Labov, 1984: 251; emphasis added), and the differential social meaning of the phenomenon built on the pertinent expression also disappears. From the other view, the concept of 'overcoming' also works when we describe or apply the concept of 'catastrophe'. Though we do not agree with the philosophical basis inherent in Ren‚ Thom's thought, his reflections become instrumentally useful. J.Petitot-Cocorda formulates a clear and essential summary of that modus operandi: "To understand a discontinuity process, one must place oneself in the following general situation. Let's consider a system S liable to certain number of stable states, controlled by a dynamics which operates in a space of descriptive parameters of the system (of the sort: phase space) called internal space. We furthermore suppose that the system S depends on a control; that is to say, it is possible to operate on it controlling the value of the other parameters which vary in another space, called, by opposition, external space. So let it S be the system for the control value s, being Ss in the stable state As. If the control s is made to vary continously, it may happen that, forágiven values of that control, a variation, little as it may be, may make the system abruptly jump from one state to another. Thus we say that there has been a crossing of a catastrophic point, or that the system has suffered a catastrophe. And in the (very frequent) event that it is the external space the one which bears the appearance of the phenomenon, its catastrophic place will be shown as a system of discontinuities that discriminate phenomenologically homogenous areas' (J. Petitot-Cocorda, 1988: 129; final emphasis added).

BIBLIOGRAPHY

(The roman numeral at the end of the references refer the reader to the notes where the text are quoted.)

(The author has traslated the quoted text that in referenced issue are not in English)

ACHINSTEIN, Peter (1968). Concepts of Science. A Philosophical Analysis. Baltimore: The John Hopkins Press (xxv)

ADORNO, Theodor W. et al. (1973). La disputa del positivismo en la sociología alemana. Barcelona: Grijalbo (Der Positivismusstreit in der deutschen Soziologie. Berlin: H. Luchterhand Verlag, 1969) (xli)

ALEXANDRESCU, Sorin (1979). La critique littéraire: métadiscours et théorie de l'explication, in A.J. Greimas, E. Landowski et al.: 208-237 (viii)

ARNHEIM, Rudolf (1971). El pensamiento visual. Buenos Aires: Eudeba (Visual Thinking. Berkley: University of California Press, 1969) (v)

ASOCIACION ESPAÑOLA DE SEMIOTICA (1986). Investigaciones Semióticas. Sevilla: Consejo Superior de Investigaciones Científicas (iv)

BACHELARD, Gaston (1972). La formación del espíritu científico. Buenos Aires: Siglo XXI (La formation de l'esprit scientifique. Paris: Vrin, 1947) (x)

BENVENISTE, Émile (1966) Problèmes de linguistique générale, 1. Paris: Gallimard (xv). (1974) Problèmes de linguistique générale, 2. Paris: Gallimard (xvii)

BERNSTEIN, Richard J. (1979). Praxis y acción. Madrid: Alianza (Praxis and Action. University of Pensylvania Press, 1971) (xxx)

BERTALANFFY, Ludwig von (1976). Teoría general de los sistemas. Madrid: FCE (General System Theory. Foundations, Development, Applications. New York: G.Braziller, 1968) (xviii)

BETH, Evert W. (1978). Implicación semántica y derivabilidad formal. México: UNAM (Koninklijke Nederlandse Akademie Van Wetenschappen, 1955) (xxxvi)

BLANCHÉ, Robert (1965) La axiomática. México: UNAM (L'Axiomatique. Paris: PUF, 1959) (xxxv)

BLOOMFIELD, Leonard (1973). Aspectos lingüísticos de la ciencia. Madrid: Taller de Ediciones (Linguitic Aspects of Science, in International Encyclopedia of Unified Science: 215-278. Chicago: The University of Chicago Press, 1938) (xii)

BOHM, David (1974). Science as Perception-Communication, in F. Suppe (ed.): 374-391 (x)

BORILLO, Michel (1977). Raisonner, calculer, in M. Borillo et al. (eds.): 1-31 (xiv)(xxxvii)

BORILLO, Michel et al. (eds.) (1977) Raisonnement et méthodes mathématiques en archéologie. Paris: CNRS

BOYD, Richard (1979). Metaphor and Theory Change: What is 'Metaphor' a Metaphor for?, in A. Ortony (ed.): 356-408 (x) (xxiv)

BRÉAL, Michel (1924). Essais de sémantique. Science des significations. Paris: Hachette (1897) (xxxii)

BRODY, Baruch A. (ed.) (1970) Readings in the Philosophy of Science. Englewood Cliffs: Prentice-Hall (viii)

BUB, Jeffrey (1974). Reply to Professor Causey, in F. Suppe (ed.): 402-409 (ii)

BUCKLEY, Walter (1978). La epistemología vista a través de la teoría de sistemas, in G.J. Klir (ed.): 219-236 (xxi)

BUNGE, Mario (1965). Intuición y ciencia. Buenos Aires: Eudeba (Intuition and Science. New Jersey: Prentice Hall, 1962) (xxiv) (xxxiv) (1973) La investigación científica. Su estrategia y su filosofía. Barcelona: Ariel (1969) (viii)

COSSIO, Carlos (1954) Teoría de la verdad jurídica. Buenos Aires: Losada (xxvi)

COURTÉS, Joseph (1980) Introducción a la semiótica narrativa y discursiva. Buenos Aires: Hachette (Introduction á la sémiotique narrative et discursive. Paris: Hachette 1976) (xxv)

COURTINE, Jean-Jacques (1981). Quelques problèmes théoriques et méthodologiques en analyse du discours, in Langages 62: 9-128 (xlix)

CUENA, José (1985). Lógica informática. Madrid: Alianza (xxiv) (xxxvii)

CURRY, Haskell B. y Robert FEYS (1967). Lógica combinatoria. Madrid: Tecnos (Combinatory Logic. North-Holland, 1958)(xiii) (xxxiii)

DESCLÈS, J.P. - Z. GUENTCHEVA DESCLÈS (1977) Métalangue, métalangage, métalinguistique. Urbino: C.I. de S. e di L. (xiii) (xix)

FEYERABEND, Paul K. (1970). How to be a good empiricist - A plea for tolerance in matters epistemological, in B.A. Brody (ed.): 319-342 (viii). (1974) Contra el método. Esquema de una teoría anarquista del conocimiento. Barcelona: Ariel (Against Method: Outline of an Anarchistic Theory of Knowledge, 17-130 in Minnesota Studies in the Philosophy of Science. Minneapolis: University of Minnesota,1970) (xxx)

FOUCAULT, Michel (1964). Sur l'interpretation, in Cahiers de Royaumont (viii) (1966). Les mots et les choses, une archéologie des sciences humaines. Paris: Gallimard, 1966) (i). (1969). L'archéologie du savoir. Paris: Gallimard (xvii). (1971). L'ordre du discours. Paris: Gallimard (xxvi)

GADET, Fraçoise et Michel PECHEUX (1981). La langue introuvable. Paris: Maspéro

GARDIN, Jean-Claude et al. (1987ª). La logique du plausible. Paris: Maison des Sciences de l'Homme (xii) (xx) (xxvii) (xliii) (xliv). (1987b). Systèmes experts et sciences humaines. Paris: Eyrolles (xiii) (xv) (xliii)

GREIMAS, Algirdas J., Eric LANDOWSKI et al. (1979). Introduction a l'analyse du discours en sciences sociales. Paris: Hachette (viii)

HABERMAS, Jürgen (1988). Ensayos políticos. Barcelona: Península (Kleine politische Schriften I-IV. Frankfurt: Suhrkamp Verlag, 1981; Die neue Unübersichtlichkeit. Kleine politische Schriften V. Frankfurt: Suhrkamp Verlag, 1985) (xlv). (1989). Teoría de la acción comunicativa, I: Racionalidad de la acción y racionalización social. Buenos Aires: Taurus (Theorie des kommunikativen Handelns. Band 1: Handlungsrationalitaat und gesellschaftliche Rationalisierung. Frankfurt: Suhrkamp Verlag, 1981) (xxiv)

HANSON, Norwood Russell (1977). Patrones de descubrimiento. Observación y explicación. Madrid: Alianza (Paterns of Discovery. An Inquiry into the Conceptual Foundations of Science. Cambridge: Cambridge University Press, 1958; Observation and Explanation: A Guide to Philosophy of Science. New York: Harper & Row, 1971) (ix) (xiv)

HATCHER, William S. (1968). Foundations of the Mathematics. Philadelphia: Saunders (ix) (xxvii)

HENRY, Paul (1986). 'On ne remplace par le cerveau par une machine': un débat mal angagé, in J.-L. Le Moigne (ed.): 297-330 (xxiii)

HILBERT, David y W. ACKERMAN (1962). Elementos de lógica teórica. Madrid: Tecnos (Grundzüge der theoretischen Logik. Berlin: Springer Verlag, 4 ed., 1958) (xxix)

HJELMSLEV, Louis (1963). Prolegomena to a Theory of Language. Madison: The University of Wisconsin Press (Editions Ejner Munksgaard, 1943) (xiii)

JACKENDOFF, Ray (1987). On Beyond Zebra: The relation of linguistic and visual information, in Cognition 26: 89-114 (x)

JANIK, Allan y Stephen TOULMIN (1974). La Viena de Wittgenstein. Madrid: Taurus (Wittgenstein's Vienna. New York: Simon and Schuster, 1973) (vi)

KANT, Emmanuel (1950). Crítica de la razón pura y Prolegómenos a toda metafísica futura. Buenos Aires: El Ateneo (Kritik der reinen Vernunft. 1781) (iii)

KARSZ, Saúl (1971). Hegel y la dialéctica, in Papeles 13: 19-34 (xxvi)

KELSEN, Hans (1956) Teoría general del derecho y del estado. México: UNAM (1944) (xxvi)

KLIR, George J. (ed.) (1978). Tendencias en la teoría general de sistemas. Madrid: Alianza (Trends in General Systems Theory. New York: John Wiley & Sons, 1972) (xxi)

KRIPKE, Saul (1980). Naming and Necessity. Cambridge: Harvard University Press (1972) (xxxviii) (xlviii)

KUHN, Thomas S. (1979). Metaphors in Science, in A. Ortony (ed.): 409-419 (x)

LABOV, William (1984). Sociolinguistic Patterns. Philadelphia: University of Pensylvania Press (1972) (li)

LADRIÈRE, Jean (1969). Limitaciones internas de los formalismos. Madrid: Tecnos (Les limitations internes des formalismes. Leuven: Nauwelaerts, 1958) (xxix) (xxxiii)

LAKATOS, Imre (1981). Matemáticas, ciencia y epistemología. Madrid: Alianza (Mathematics, Science and Epistemology - Philosophical Papers. Volume 2. Cambridge: Cambridge University Press, 1978) (l)

LECOMTE, Alain et Pierre PLANTE (1986). Espace des séquences; approche topologique et informatique de la séquence, in Langage 81: 91-109 (xvii)

LEFEBVRE, Henri (1984). Lógica formal, lógica dialéctica. Madrid: Siglo XXI (Logique formel, logique dialectique. Paris: Anthropos, 2nd printing, l969/ manuscript from 1947) (li)

LE MOIGNE, Jean-Louis (1986). Genèse de quelques nouvelles sciences: de l'intelligence artificielle aux sciences de la cognition, in J.-L. Le Moigne (ed.): 15-54 (vi) (xlii)

LE MOIGNE, Jean-Louis (ed.) (1986). Intelligence des mécanismes, mécanismes de l'intelligence. Paris: Fayard (xxiv)

LÉVI-STRAUSS, Claude (dir.) (1988). La identidad. Seminario. Barcelona: Petrel (li)

LYONS, John (1977). Semantics, I, II. New York: Cambridge University Press (xv)

MAGARIÑOS DE MORENTIN, Juan A. (1986). Logic Foundations of Semiotics, in Semiotische Berichte 10: 141-159 (xvii) (xlvii) (xlviii) (xlix). (1992) La recurrencia semiótica, in Signa 2: 105-111. Madrid (ix). (Forthcoming) Comentarios metodológicos a 'Fundamentos Lógicos de la Semiótica'. La Plata: IICS (xli) (l)

MAGARIÑOS DE MORENTIN, Juan A. y colaboradores (1993). Análisis Semiótico del Discurso Político. La Plata: IICS (1989) (xxii)

MALMBERG, Bertil (1969). Lingüística estructural y comunicación humana. Madrid: Gredos (Structural Linguistics and Human Communication. An Introduction into the Mechanism of Language and the Methodology of Linguistics. Berlin: Springer Verlag, 1967) (xviii)

MARANDIN, Jean-Marie (1979). Problèmes d'analyse du discours, in Langages 55: 17-88 (xii)

MARR, David (1982). Vision. New York: W.H. Freeman (xxv)

MARTY, Robert (1986). La semiótica de C.S. Peirce: presentación formalizada, in Investigaciones Semióticas I: 339-353 (iv)

MEUNIER, Jean-Guy (1989). Artificial Intelligence and Sign Theory, in Semiotica 77: 43-63 (xxi)

MONTAGUE, Richard (1977). Ensayos de filosofía formal. Madrid: Alianza (Formal Philosophy - Selected Papers of R. M. New Haven: Yale University Press, 1974) (xviii)

MOORE, George (1974). Defensa del sentido común, in J. Muguerza (ed.): 252-277 (A Defense of Common Sense, in Contemporary British Philosophy. London: G. Allen & Unwin, 1925) (xxxiii)

MORRIS, Charles (1969). Foundations of the Theory of Signs, in O. Neurath, R. Carnap, C. Morris (eds.): 78-137 (xii)

MUGUERZA, Javier (ed.) (1974). La concepción analítica de la filosofía, 1. Madrid: Alianza (xxxiii)

NAGEL, Ernst y James R. NEWMAN (1970). El teorema de Gödel. Madrid: Tecnos (Gödel's Proof. New York University Press, 1958) (xxxiii)

NEURATH, Otto; Rudolf CARNAP; Charles MORRIS (eds.) (1969). Foundations of the Unity of Science. Chicago: The University of Chicago Press (1938) (xii)

ORTONY, Andrew (ed.) (1979). Metaphor and Thought. New York: Cambridge University Press (x)

PANIER, Louis (1979). Le discours d'interpretation dans le commentaire biblique, in A.J. Greimas, E. Landowski et al.: 239-254 (viii)

PARRET, Herman (1983). Semiotics and Pragmatics. Amsterdam/Philadelphia: John Benjamins (xii) (xxiv) (xlviii)

PECHEUX, Michel (1975). Les vérités de La Palice. Paris: Maspéro (iii) (xii)

PEIRCE, Charles Sanders (1965). Collected Papers. Cambridge: The Belknap Press of Harvard University Press (1931) (iii) (vi) (vii) (ix) (xvii) (xxiii) (xlv)

PETITOT-COCORDA, Jean (1988). Identidad y catástrofes (topología de la diferencia), in Cl. Lévi-Strauss (dir.): 119-173 (li)

PICO, Josep (ed.) (1988). Modernidad y postmodernidad. Madrid: Alianza (xxxii)

POPPER, Karl R. (1977). La lógica de la investigación científica. Madrid: Tecnos (The Logic of Scientific Discovery. London: Hutchinson, 1958; 1st german printing: 1934) (xxviii)

RASTIER, François (1987ª) Présentation, in Langages 87: 5-19 (xxiii). (1987b) Représentation du contenu lexical et formalismes de l'intelligence artificielle, in Langages 87: 79-102 (xlvi)

SALLANTIN, Jean (1986). Logiques et comportements des systèmes rationels, in J.-L. Le Moigne (ed.): 117-137 (xv) (xlvi)

SAUSSURE, Ferdinand de (1972). Cours de Linguistique Générale. Paris: Payot (xvii)

SHAPERE, Duley (1974). Scientific Theories and Their Domains, in F. Suppe (ed.): 518-565 (xix) (xliv)

STURTEVANT, E. (1947). An Introduction To Linguistic Science. New Haven: Yale University Press (li)

SUPPE, Frederick (ed.) (1974). The Structure of Scientific Theories. Chicago: University of Illinois Press (ii) (x) (xiv) (xv) (xix) (xliv)

SUPPE, Frederick (1974). The Search for Philopophic Understanding of Scientific Theories, in F. Suppe (ed.): 1-241 (xv)

SUPPES, Patrick (1974). The Structure of Theories and the Analysis of Data, in F. Suppe (ed.): 266-283 (xiv)

VOLPE, Galvano della (1966). Crítica del gusto. Barcelona: Seix-Barral (Feltrinelli, 1963) (xxxix)

TARSKI, Alfred (1956). Logic, Semantics, Metamathematics. Papers from 1923 to 1938. Oxford: Clarendon Press (xiii)

THOMASON, Richmond H. (1977). Introducción, 9-77 in R. Montague (viii)

WELLMER, Albrecht (1988). La dialéctica de la modernidad y postmodernidad, 103-140 in J. Picó: (On the dialectic of Modernism and Postmodernism, in Praxis International 4/4, 1958) (xxxii)

WRIGHT, Georg Henrik von (1970ª).Ensayo de lógica modal. Buenos Aires: Santiago Rueda (North Holland, 1951) (xxvi) (xxviii). (1970b). Norma y acción. Una investigación lógica. Madrid: Tecnos (Norm and Action. A Logical Enquiry. London: Routledge & Kegan Paul, 1963) (xxvi)

ZELENY, Jindrich (1982). Dialéctica y conocimiento. Madrid: Cátedra (Studien Über Dialektik, in Filosoficky casopis, 1972, 1973, 1974) (xxxii)

ZIMA, Peter V. (1989). Parataxis und Dekonstruktion. Von Adorno zu Derrida, in Semiotische Berichte 13: 4-17 (xxxix) (xliv)

__________

PAGINA PRINCIPAL