Doing things with words vs. Saying things with data

Today it seems redundant to say that words have an active function in the way we think or speak (since much has been said about it). But in 1955, when John Langshaw Austin, at Harvard, said that words not only serve to say things but that they “do”, he changed the way we conceive discourse.

His position is strongly influenced by language thinkers such as Benveniste, Wittgenstein, Malinowski, and Jakobson. This problem starts from the premise that words have significant consequences in components that are not strictly linguistic. The making power of words exceeds their verbal limits.

But, beyond the theoretical, something else worries me: what happens when words and, specifically, language, are placed at the service of the market (specifically, the digital market)? In other words, what happens when it is given a “useful” function? Here we begin to think that many of the things we say are said because someone (or something) led us to tell them.

Data analytics, databases, machine learning, digital data, and language processing technologies have something to say on the matter. This could be where the problem of “doing things with words” today leads to “saying things with data”. 

In this article, I’d like to address this problem. I will talk about how interesting Austin’s proposal is, but, in addition, I will mention how this situation evolves and becomes more complicated when the language is subscribed to the digital market.

What John Austin says to “doing things with words”.

Austin in 1955 presented his Harvard University lectures on How to do Things with Words. He presents with a curious idea: words can create, form, make, say, and build situations.

Later, in 1962, the book How to Do Things with Words was released. In this work, Austin introduces and develops the theory of speech acts. It examines how words serve not only to describe the world but also to chain actions. In a nutshell, the book deals with the problems of:

• Distinguishing realizational utterances. Distinguishing between constatative utterances, which describe the state of affairs (whether true or false), and realizative (or performative) utterances. 
• Speech acts. One of the best-known points of his lectures. He identifies three types of speech acts, which would be, the power and effects of speech. They can be emotional/psychological/linguistic acts of words, which are presented as locutionary acts, or saying something with a certain meaning and reference; illocutionary acts, which are performed by saying something (such as promising, ordering, asking, etc.); perlocutionary acts, or the effects or consequences that words have on the listeners (such as convincing, frightening, inspiring, etc.).
• Conditions of success. Certain felicity conditions must be met for an illocutionary act to be successful. For example, in making a promise, one must intend to keep it and be in a position to do so.
• Felicity and failure of speech acts. Discusses how speech acts can fail. For example, someone promises something without the intention of fulfilling it. Then, the promise is unhappy.
• Importance of context. Emphasizes the importance of the context in which speech acts are performed. The meaning and effectiveness of words depend to a large extent on the circumstances in which they are uttered.

How “doing things with words” is complicated on the Internet

We advance to the second level of the problem. Words have a power and a consequence. They can be used for what they say and the possibility of situations they generate. But, what happens when that word that “does” is also capable of materializing/functioning? At this point, I want to get into the subject of the internet and digital media.

The things we say today are tremendously conditioned by the information we receive. This information, moreover, is filtered by algorithms. In other words, we are talking about codifications built from quantitative languages that generate responses and actions based on commands. 

Media buying campaigns and different marketing strategies filter certain types of content. Nowadays, doing things with words has a strong implication on consumer interests, of sellers and buyers. The language that “does” also “says”, but is backed by data. It is data that organizes, quantifies, categorizes, and prioritizes certain types of content.

As Boris Groys mentions in his ideas of Art in Flux, the graveyard of web pages created by web browsers is the consequence of a deliberate manipulation of digital language. In this context, language plays an active (but automated) role. It behaves as a referent. Language is subjugated to the analytical role of a judge, which, automatically, determines what is being used and what is not.

Saying things with data to sell

Saying things with data means that words that describe things are only there because a digital language led us to it. The content we consume, or even the way we think, is involved by the media. 

Whether we sit down to read a book, watch the news, enjoy a football match, or play video games, all these acts were deliberately mediatized for us to consume them. And who mediatized us to do so? It may be a private company or a government party, but language is what brought about this phenomenon.

Computer languages are supported by data analytics, databases, and other marketing strategies so that these conditions are applied to our consumption style. A market superstructure has turned language into a tool to conceive us as pieces of human capital.

Doing things with words that sell

Behind the feeling that we are masters of our decisions and that the world is full of things to choose from, we find ourselves trapped inside a wheel that crudely makes us repeat and do the same thing permanently to safeguard the capitalist system. In this case, digital languages are promoting this situation.

Austin’s proposal in his lectures extends to a performing function. But, taken to a modern context, we see that the word (concretely, language) speaks that not only to say things is to think a world but that to think of things is itself to make, to produce. 

The deliberate linking of language by the interests of some sectors is a way of devaluing its value. And, at the same time, the productive power of language as a mechanism of syntax is thereby fully understood. Here companies, in the backstage of the backend of a server, tell us what to say, when, and how, without offering us a why. They use the language to generate this event.

I think there is a bad intention behind the use of language in digital media. Today it is a subscriber to the means of reproduction of the digital capital market. This language quantifies and enumerates what we say (or what we seek) to turn it into an object of consumption on the Internet.

The irresponsible and excessive use of these “words” as “tools”, for example, in a content campaign, can have a great impact on the development and formation of new digital technologies.