As promised in the previous episode, we need to delve deeper into the connection between language and method , which—perhaps a somewhat convoluted path—I've actually been debating since my very first contributions to this column. As I tried to explain previously, if we could establish that language is the method, we would have a powerful theoretical basis for defining the nature of data, the procedural organization of research (i.e., techniques), and criteria for interpreting the results. Therefore, not sociological theories to describe and interpret the results (which are and remain fundamental), but a sort of meta-theory , a theory of method that precedes and guides every subsequent social theory. For those who are skeptical, I'll begin by showing how the semantic level is already widely considered in methodology, and I'll begin with a great classic, the Lazarsfeldian paradigm.
What is the Lazarsfeldian paradigm?
The Lazarsfeldian paradigm is nothing more than a logical trace, a general enough path to guide the researcher in the breakdown of a complex concept in order to identify the indicatorsA step back: when we begin a new research project, the subject usually refers to broad, multifaceted, multidimensional issues, such as: "The crisis of the family in post-industrial society"; "Young graduates and Industry 4.0"; "The diffusion of stereotypes in globalized society." Each of these three titles, these three research objects, can be defined as concepts that – following marradi – can be described as “places of socially constituted meanings, which contribute to the organization and communication between individuals: 'table' is a concept, like 'anger', like the sentence 'Voters decide at the last moment which party to vote for', which is a complex concept made up of several concepts.i”. Now: it is evident that a concept (the object of our hypothetical research) such as “The crisis of the family in post-industrial society” cannot be operationalized in any way given its great abstractness; what type of 'crisis' are we talking about? Which 'families'? What do we mean by 'post-industrial society'? The researcher must therefore better define each of these specific elements, breaking down their meaning into constituent elements lower on the scale of generality. The Lazarsfeldian paradigm is nothing other than the general scheme which is followed in this descent, breaking down an abstract concept into dimensions (still quite abstract and high on the scale of generality), then possibly into sub-dimensions (less abstract, less “high”), and finally into indicators.
The level of indicators
If this seems too far-fetched, here's the same scheme applied to a hypothetical survey of drug addiction. The final level, called the indicator level, is the level for which the researcher—due to its low level of generality—can imagine operational definitions, or empirical methods of data collection (for example, questions in a questionnaire). Let's return to semantics: this entire process is based on a semantic relationship between concept, dimensions e indicatorsEach indicator – it is assumed – semantically “covers” a portion of the dimension which, analogously, semantically covers a portion of the concept.
Semantic coverage
This semantic coverage It is not measurable; it cannot be determined, validated, or certified in any way. It has a semantic nature, which means that only an argumentative framework can demonstrate (not "prove") that those dimensions, and not others, are appropriate, suitable, and complete, and that therefore the concept is fully covered (if this were not the case, the uncovered parts would not become the object of analysis, resulting in a clear cognitive deficit in our research) and that other, extraneous concepts have not been covered – by mistake (this would actually be worse, because we would be exploring something else, perhaps similar but different, without realizing it). The same goes for indicators. The operational definitions of a research project, in essence, should be the result of a semantic process of this kind, and, having this nature, the operationalized indicators are subject to the rules of semantics and its logic and not, for example, to hypothetical and unrealistic causal logic, if anything determined and "validated" by indices of some kind. To clarify, I propose a third figure that shows what is meant by “semantic coverage” of a concept (in this case 'youth distress').
Consider this brief presentation of the Lazarsfeldian paradigm a small introduction to semantic logic applied to social research. Next time, I'll show you how semantic reflection guides the management of work groups.
Claudio Bezzi
