R: A Language Born to Think with Data

What does programming have to do with reading, interpreting, and producing knowledge in social sciences and humanities? This text proposes an entry into R from its history and academic uses, showing how its design, its function-based logic, and its package ecosystem interact with familiar analytical practices. More than a technical introduction, it is an invitation to read code as language and to think of programming as a situated research practice.

When we, from the social sciences and humanities, first approach a programming language, an ambiguous feeling often arises: curiosity, interest, and, at the same time, a certain caution. Programming is frequently associated with a technical practice, distant from reading, discursive analysis, or theoretical reflection. However, we can choose a different entry point to learn R. Its history, its uses, and the community that sustains it make it a language particularly close to an analytical perspective on data and on the conditions of knowledge production.

R originated in the field of academic statistics. In the early 1990s, Ross Ihaka and Robert Gentleman, statisticians at the University of Auckland, developed R as a free re-implementation of the S language, used since the 1970s in statistical research. From that initial moment, R was conceived as a tool for exploring data, testing models, and supporting scientific research processes. This origin helps to understand many of its current characteristics and also the practices that have consolidated around its use.

One of R’s distinctive features is its orientation towards interactive analysis. The code executes progressively and allows for exploratory work with data: trying an operation, observing the result, adjusting the instruction, and re-evaluating. For those accustomed to building analyses iteratively, based on provisional hypotheses and constant reformulations, this mode of work is particularly familiar. Programming in R usually involves sustaining a process of controlled exploration, rather than following a completely predefined path.

From a technical point of view, R relies heavily on the use of functions. Most operations are performed by applying functions to objects (in fact, R is an object-oriented language, just like Python). A function can be understood as a formalized operation that receives one or more arguments, executes a series of internal steps, and returns a result. Calculating an average, transforming a variable, fitting a model, or producing a graph are examples of tasks performed using functions. Reading code in R involves identifying which functions appear, what arguments they receive, and what type of object they generate as output.

This emphasis on functions favors a careful reading of code as if it were a text. Each function call presents a relatively stable structure: a name, parentheses, and arguments that can be explicitly named or implicitly defined by default values. For those with experience in text analysis, this syntax can be approached as a system of markers that guides interpretation. Understanding why a function produces a certain result often requires pausing on these details and on the assumptions that the function incorporates.

Over time, the uses of R expanded considerably. Currently, it is used for statistical analysis, modeling, data visualization, text analysis, social network studies, working with spatial data, and experiments in psychology and linguistics, among many other fields. This expansion is largely explained by the package system. A package is an organized collection of functions, data, and documentation that extends the language’s capabilities. R thus functions as a constantly growing environment, fed by contributions from different academic communities. For the social sciences and humanities, this feature is particularly relevant. Many of the tools available in R were developed by researchers working on problems similar to ours. Packages oriented towards the analysis of surveys, texts, linguistic corpora, or longitudinal data incorporate specific theoretical and methodological decisions. By reading the code that uses these packages, one also accesses these decisions: how a unit of analysis is defined, which operations are considered relevant, what assumptions are adopted about the data.

In this sense, R can be thought of as a language in which particular ways of thinking and researching are inscribed. The existence of object classes, the representation of models as examinable objects, or the construction of layered graphics organize the relationship between the analyst and the analyzed phenomenon. Learning R implies learning to recognize and work with these mediations.

For this reason, in a series aimed at those beginning programming from the social sciences and humanities, it is productive to present R as a language with a history, conventions, and communities of use. Reading code in R, even when it initially takes time and effort, is part of a technical literacy that dialogues with habitual practices of reading and interpretation.

In the next texts in this series, we will introduce concrete code snippets. The goal will be to learn to read them, decompose them, and understand what they are doing and what they are saying. Programming in R, from within our disciplines, implies incorporating a new form of writing and of situated knowledge production.