Abstract
This article presents two lines of thought that provoke tension among some postulates related to legitimacy in qualitative research; those related to issues like validity as a way to achieve legitimacy. I take a stance for research, basically, as an activity of discursive production that follows argumentation to guarantee its legitimacy. Finding support in some key ideas from Roland Barthes, Jesús Ibáñez and Irene Vasilachis, I depict an epistemological perspective that falls within the current discussion about qualitative research in social science.

Key words author
Epistemology, second rate research, research legitimacy, objectivity, discursive education.

Key words plus
Qualitative methods (research), theory of knowledge, information theory in the social sciences.

Transference to practice
Although, these ideas belong to epistemological reflections, they can contribute to rethink the research work of undergraduate and graduate students that face difficulties while constructing their subjects and objects of research.

In empiric research, what we usually call data, are actually capta: since they are selected arbitrarily, because the form of the data depends on the reference framework and this reference framework is a function of the distinctions and indications of the researcher.
Jesús Ibáñez (1998, p. 24)

What is certain is that only God can tell infallible humans the “real” nature of reality.
Strauss & Corbin (2002, p. 5)

Introduction

According to Barthes (1987), to positivist science, language is the means to deliver its messages. Those messages, science would say, are outside language, because its main concern is the facts, not the words that tell about those facts. On the other hand, and the other way around, the essence of literature is language, words, not facts, since “language is the self of literature” (Barthes, 1987, p. 15). The way of saying things is its field of work that is its real concern, not the referent of what was said: the message, the contents, the researched reality, and the findings.

Because of what I said above, and from that framework, I insist science, positive science, is not inside language, as Barthes (1987) would say, because science is what is said: the studied phenomena. However, if the essence of literature is language, its main concern will be how is said, and its main concern will be the writing, the text. To us, the ones who research, and try to write, it is clear: writing is hard work; it is a battle, a struggle against signifiers (Zuleta, 1994), because writing is power, a field of power: multiple voices speak in the text, not always identified by us, not always controlled by us, and they speak in particular ways. We are spoken in our texts. In the text, we also know it, the voices we silenced, when choosing how to write our text, breath.

From this angle, the text, the research article, the report, could be seen as the field of power, like the marks that tell about that struggle with words, with the data, because writing contains the trace of practice, in our case, research practice. Maybe, this is the reason why Irene Vasilachis (2007) explains it is better to talk about epistemological reflection than epistemology, in her argument about the premises of qualitative research; there she proposed an epistemology of the known subject. This epistemological reflection is a key element in the researcher’s work; it works as a kind of intellectual alarm to uncover naivety.

On the perspective of building a point of view about research, precisely qualitative research in social science, from a non-positivist horizon, these lines of thought turn out to be very suggestive, if we invert the order of the vectors, as we have outlined. Firstly, because research could be conceived beyond its findings and the facts it tries to describe, explain, or prove, it could be seen as a way to say, as a language. Secondly, because the requirement of rigor, from this perspective, should be looked for within the system of words, the conceptual system or the body of categories used to talk about [the research], in the research. Its meaning, or meanings, could be found if we analyze this particular way to say, the text, that is to say, we could not look for them only in the facts that supposedly these words account for, nor in its findings. In this respect, the legitimacy of research, the search for recognition in the academic community and in society, in general, could not be looked for solely in its results, neither in its applicability, nor in its transference to practice, nor in its incidence in life and policy decision-making. Let us not forget that history shows us all the time the complexity that surrounds the diverse and uncontrollable use of the results of research, something researchers can barely calculate or dimension and that in many occasions illegitimates research.

But, what if we consider science, social research, a way to speak, we will curiously be getting closer to literature, its strength, its value, its legitimacy, its validity, must be looked for in that particular way to speak, and not in the correspondence of that way to speak and the facts it mentions, wants to mention or says to represent. This value should be looked for in the way the statements that speak of –and in– the research were chosen, ordered, produced. This view would demand to ask ourselves beyond what the text wants to mention, for the situation in which the statements were produced: whose side is the statement taking?, since they will never be neutral, whose voices are speaking through them?, whose voices were silenced o excluded?, what was left in the interstices, in the ruptures, in the discontinuities, that are always there, that are hardly hidden?

The impossibility of representation

In the positivist and pospositivist tradition, even in some contemporary qualitative studies, legitimacy is usually looked for, at least partially, in validity, but the problem of validity in social science research is still unresolved. Positivist science presents validity from the connection among results, explanations, descriptions or hypothesis and the reality, the situation, the facts or phenomena those explanations deal with. The problem with this notion of science, if we follow Barthes, would be that they are still presenting an outside of language, a dichotomy. This version of science would keep on putting forward the existence of what is real, outside the words that mention it, something unacceptable according to Barthes (1986), given the impossibility of representation of language, language does not represent: it expresses a reality, shapes it. From this viewpoint, we can affirm that when we research, and mention the phenomena we are studying, we promote their existence, and link it to the discursive order we are following, the system of decisions and positions we assume as academic and political subjects.

From this line of argumentation and analysis of research, coming from two different traditions and disciplines, that I find very similar: Barthes (1986, 1987) with his reflections about language and text; and Ibáñez (1994), with her epistemology of social science for a second rate research; it is feasible to affirm that since language is a classification, what it mentions, the researched object, is transformed depending on the system of concepts and instruments used. This system is usually called categories, conceptual system, and analysis matrix. That is why, apart from the reasons given above, it would not be possible to claim the existence of the facts researched as external to the language that mentions them or independent from the instruments that shape them. Thus, the facts, the social phenomena, would be, to a great extent, an effect of the language they are said in, produced in. That is why, it does not sound strange to accept that social reality is a construction: those words are the title of a rather classic play. From this particular theoretical frame, validity, defended from the relation between the explanation and the fact, guaranteed by the control of the instruments and procedures to reduce bias, and the fact mentioned, would be impossible from an epistemological point of view. This line of argumentation, makes us, or at least authorizes us to search for legitimacy in other grounds, for example in the borders of a discursive, argumentative perspective.

It is possible to indicate that this line of thought is related to the developments of Michel Foucault, in his analysis of the way the objects of study of scientific disciplines have constituted, since, in them,

If there is unity, the principle is not, then, a determined way of statements; wouldn’t it be instead the set of rules that have done, simultaneously and consistently, possible purely perceptive descriptions, but also observations influenced by instruments, protocols of laboratory experience, statistic calculations, epidemiologic or demographic testing, institutional regulations, therapeutic prescriptions? (Foucault, 2007, p. 56).

That is to say, scientific disciplines can be considered like discursive constructions derived from the way statements have been chosen and organized, that is, based on the existence of certain instruments that organize the discourse and that, at the same time, depend on historic situations. Thus, following Foucault, the supposed stability of a discipline, of a science, does not obey to the stability of an object of study, since this object is, to a great extent, produced by those instruments, by those statements. Again, that supposed discipline’s stability is fundamentally controlled by a discursive order.

And then, from where is the legitimacy of a research postulated?

From the ideas we have exposed, the legitimacy of a research could be postulated from the internal architecture of the system of concepts, instruments and words that the researcher builds to mention, describe or to explain his/her research object or social phenomena, something that could only be taken as a discursive construction. In this perspective, the researcher as well as the research will be asked to be clear, to be explicit with regards the position they are speaking from; to explain what operations were made to build their discourse; why they chose those words, those concepts and not others. Like Vasilachis (2007, p. 30) states “the analyst has the obligation of checking and presenting her/his own analytical process and procedures as complete and truthful as possible”.

In this framework, the recognition of a research and discursive practice could take the path of a discursive, textual, grammatical analysis. If there is a thorough discourse analysis that gives account of the research, we will obtain helpful clues for the adjustment and eventual transformation of the research practice since the transformations worked on language, on the concept system, could point out eventual transformations on the experience, as far as they give some insights on its operation, its limits, and its strength. To sum up, the task of decomposition and composition of the language used by the researcher would give an account not only for the structure of the discourse built but also for the research practice, like Barthes (1987) says, a text is the spelling of practice. Inside and in the borders of the text, the decisions of the researcher are breathing, even in what is not said in the writing; just as in the style rest the traces of a struggle with words, again, also with the data which is produced by that way to say. And then we come again to the eternal circle of interpretation. Apart from what we said above, there is another request, another question to any research that postulated itself as neutral, and to the researcher: what are the operations and conditions which help the discourse let go of power? (Barthes, 1986); since, speaking, at least in the academic and the research environment and given the social status of this kind of knowledge, is, inevitably, to have power. Facing this, and from another theoretical place, we would say that when we speak our minds, we take sides (Anscrombe & Ducrot, 1983), and this assertion makes us ask ourselves, what kind of power do we want to exert with our way to say?, or what kind of power is being exerted on our way to say?, what kind of power is speaking through our voice, our text, our research?

To conclude, from this viewpoint, research does not describe or present reality: it produces it. That production is rooted in language, since the form of the reality produced depends on the system of words chosen for that speaking. When speaking, we produce reality, that product depends on the election of the system of words and the ways they relate. Thus, and from this perspective, just a perspective, just like literature, in research we create new worlds, symbolic objects, because from the very first moment we decide on a research, when we formulate a clear research question, a path is open. From there, from those first decisions, we define what data we will take but, actually, the data will be an effect of those first decisions because they will be limited by the words and instruments we choose. And when we describe, when we make observations about social phenomena, from those decisions, the observations become new data, and we go back to them to construct the next interpretation, again, and again and again, until the next report, until the next article. Then we go back to the interpretation and keep going on an infinite and consecutive semiosis, since this is a symbolic production, which comes from another symbolic front. (Ibáñez, 1998).

Notice that in this way, the legitimacy of research will not be postulated only from methods, techniques and the possibilities of transference of knowledge, what we call impact or objectivity in the results. Besides, and to add another angle for the analysis of this problem, that we will not develop here, we know that qualitative research is controlled by diversity: of methods, of approaches, of epistemological positions of methodological constructions. Therefore, reducing the research phenomenon to, for example, a problem related to the election of a method is untenable. Thus, in a way, the results of a research process can be seen as the subproduct of the process of discursive construction. That is why the argumentation and the search for discursive consistency are key elements of social research.

Of course, many questions are left unresolved: what differences can we postulate in the way to say we call research and in the other ways to say?, since we are talking about a serious, academic speaking, it will be legitimated if it states explicitly the place it is speaking from. The researcher will be asked to explain why she/he chose those and not other words to express ethic and political positions. She/he will be asked in whose favor her/his research is speaking to. In short, argumentation will be the only possible setting for all explanations.

Our hope is that these ideas contribute to deliberation on the complex task we face: researching and writing. That is what I meant in this writing: to contribute to a potential discussion.

References
Anscombre, J. C. & Ducrot, O. (1983). La argumentación en la lengua. Madrid: Gredos.
Barthes, R. (1986). El placer del texto y lección inaugural. México: Siglo XXI.
Barhes, R. (1987). El susurro del lenguaje. Barcelona: Paidós.
Foucault, M. (2007). La arqueología del saber. México: Siglo XXI Editores.
Ibáñez, J. (1994). El regreso del sujeto. La investigación social de segundo orden. Madrid: Siglo XXI Editores.
Ibáñez, J. (1998). Nuevos avances en la investigación social: La investigación social de segundo orden. Barcelona: Anthropos Editorial.
Strauss, A. & Corbin, J. (2002). Bases de la investigación cualitativa. Técnicas y procedimientos para desarrollar la teoría fundamentada. Medellín: Editorial Universidad de Antioquia.
Vasilachis, I. (2007). Estrategias de investigación cualitativa. Barcelona: Gedisa.
Zuleta, E. (1994). Elogio a la dificultad y otros ensayos. Cali: Fundación Estanislao Zuleta.